Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
The Internet

What Should One Look For in Colocation Services? 174

willeg asks: "I am charged with the task of recommending a company for co-locating our auction site. The problem is that on paper, they all look about the same. What other information should be used in order to make the decision on a good hosting pick service?" This will probably be a question that many will ask in the future. With the Internet being the 'next big thing', everyone will be looking to stake a claim on their own portion the digital real-estate. Problem is, not many people can afford the bandwidth costs of getting wired for high-speed access, especially for businesses in out-of-the-way areas. Is colocation the easy answer? Or is just as expensive?
This discussion has been archived. No new comments can be posted.

What Should you Look for in Colocation Services?

Comments Filter:
  • by Chiasmus_ ( 171285 ) on Monday June 19, 2000 @10:27AM (#991531) Journal
    Different services require different amounts of bandwidth. The answer to this question is going to be pretty different depending on whether you're planning on putting up one "order form" for a small business, or a giant search engine that relies heavily on banner ads. It's also going to be different depending on whether you're an IT company already and have some of the equipment and personnel available already, or you're some random industrial company that just wants a web presence.
  • First and foremost you want intelligent administrators at your colo. Ones with a structured regiment for backups and preventive maintenance. A very important thing you will see really quick is that you also must require good customer care. For some people 24 hour respond time is okay, for some you need it at that exact second. If the second is what you require i dont recommend colo at all of course. Basically it comes down to service.
  • by synx ( 29979 ) on Monday June 19, 2000 @10:29AM (#991533)
    The problem with colocation is that everyone promises the world, but non deliver.

    You want the following features:

    * electronic security (good to find out who was in when)
    * multiple 100 meg pipes...
    * full BGP4 routing so all those pipes are used at all times, not just when one fails.
    * good connections... who are they peering with?
    * indivdual locked cages, video security is very good.
    * switched ethernet (ok, if you're talking 100 meg colos... but if you find you're on a hub, thats a major sign they are incomentent)
    * obscure locations - you don't want a sign saying "very expensive computers inside"
    * 24/7 on call support if necessary... what if you need to have that critical machine rebooted at night? sure it'll cost you, but it'll cost you more to have it down.

    I work at a colo company, and our fast facility meets these requirements... Actually, I note that Vancouver is one of the best wired cities in North America, we got fat fat pipes to Seattle which is of course set up beyond belief..

    Anyways...
  • You might want to check out an excellent article [hostfind.com] at TopHosts.com [hostfind.com]. You server needs to be in a secure location, physically, so it's safe from damage (natural, ie earthquakes, as well as man-made, such as theft.) It's also best if you physically own the server, and can remove it entirely if you wish. Redundant connections to the net are also useful so that you can't easily be cut off.
  • You can search for weeks and not find what you want. There are companies that help you find co-lo, circuits, etc, by either putting you in touch with the right people, or with a larger need can run reverse auctions to get you a better price. It takes the guesswork out of it. Band-X (www.band-x.com) is the service out there right now.
  • by FascDot Killed My Pr ( 24021 ) on Monday June 19, 2000 @10:31AM (#991536)
    ...ask THEM.

    Ask all of your bidders: "What can you give me that the others can't?" You are sure to get a bunch of crap like "a sense of well-being that your blah blah blah". Take all non-nonsensical (would that be "sensical"?) responses and redistribute to the vendors for response. Let them respond again.

    At the very least, this will weed out the people who don't know what they're talking about.
    --
    Compaq dropping MAILWorks?
  • Hey, I was just looking for a colo in Vancouver,BC... email with info to sheldon_young@yahoo.com. Thanks.
  • Here's generally what I expect (in Canada mind you). $250 Canadian a month should get me a single server on a 10Mbps switched port, on my own VLAN, generally connected directly to at very least a T1, but as times are changing often an OC3 in my area.

    I like to know that whomever I'm collocating with is on a peered network (connected directly to uunet, sprint, or whomever), and that I'm getting high ping times and good traceroutes.

    Redundancy. C'mon, this is 2000. Make sure they have some good old UPS's in place, and that you're guaranteed 99.9% reliability if at all possible.

    You're never asking too much when you want your business to have minimal downtime.

    Security? It might be worth the extra few bucks for a secured room on their premises. Who has access to your box?

    My $0.02...

    EraseMe
  • Are you needink to have physical access to the computer?
    If you are then look for a place where the building is open as many odd hours as you can find.
    Looking to just leave it there and do all the work by remote?
    Look for a place that has nice people who are smart enough to cycle power for you ar do other little odd things that might need doing.
    Other than that every ont your looking at should have there co-los stored within FCC regulations.
    -----
    If my facts are wrong then tell me. I don't mind.
  • 24 Hour a day access to the co-lo site sure is nice. If your system bombs in the middle of the night, it sure is nice to be able to walk in and fix it, rather than wait until the morning after your clients have put a few hundred emails into your box...
  • by Cliffton Watermore ( 199498 ) on Monday June 19, 2000 @10:34AM (#991541) Homepage

    Ask them "How many hops to the backbone?". Ask them how much bandwidth they have to their nearest peering point. Most importantly, ask them if you can put a test server in one of their racks, FTP and TTCP to it (Test TCP), really drown the link - to see what it's capable of, and then insist on comparing the results to similar results after your contract is processed. (Don't put your Auction Box in there until you've re-tested with an identical FTP set up).

  • 1) Price (this is a no-brainer), both the up-front cost and the bandwidth charges.

    2) Connectivity: who does the host peer with and what is the quality of his connections.

    3) Quality of Service: you do not want to travel to the colocation facility in order to handle basic administration tasks. What is their customer service record and are they providing you with rudimentaty system administration or just rack space?

    4) Security: some premium web providers are popping up that offer additional security and support as value added services. If you want to colocate a mission critical system, this option might be for you--but be ready to pay through your nose for it.

    5) Reliability: what is the host's track record? Would you be comfortable trusting your business to them?
  • Vancouver, BC, or Vancouver, WA?

    One's across the border to the north; one's down by Portland, OR

    There's two of 'em: confusing as hell, on occasion..

    t_t_b
    --

  • by slashdotter168 ( 126598 ) on Monday June 19, 2000 @10:38AM (#991544)
    You also want to know:
    -what kind of heating/cooling the facility has?
    -is your equipment going to be off the floor in case of a flood?
    -what kind of fire supression equipment do they have?
    -are the NOC monkeys knowledgable?
    -do they monitor your equipment for you?
  • by DMC ( 10005 ) on Monday June 19, 2000 @10:39AM (#991545)
    a few words of advice with the first one being "STAY THE HELL AWAY FROM LEVEL3". whew. i haven't seen service this bad from a bell. all of their procedures are broken and misunderstood internally. cross-connects take a minimum of 14 business days. the sales people never call me back, and take forever to enter orders into the system. it's not like we are a one off customer either. we are in 15 of their us colocs with plans to go into the rest of their us sites as wells as several of their international sites. this plan is changing tho, and we are considering moving everything to worldcom. they may be a bit slow, but they get things done and don't pull new policies out of their ass.

    at level3 the people in the gateway are pretty good on average, but the people up the chain from there are pretty much a waste of o2.

    my experience with abovenet in san jose has been pretty good. they are helpful and fast. they are well connected and offer lots of service levels.
  • by FFFish ( 7567 ) on Monday June 19, 2000 @10:39AM (#991546) Homepage
    Once again, Kuro5hin and Slashdot are duplicating each other. This recent thread [kuro5hin.org] on Kuro5hin had some answers.

    --
  • You mean low ping times, right?

    And as for traceroutes, would you be looking to make sure you aren't hopping like 5 extra times inside their network?

    The ping times could be a simple mistake or , maybe i'm missing something. I'm interested in details you have mentioned, so any greater details woudl be nice to know.

    thanks

  • I'll just use this question to post a related one:

    I'm interested in renting (leasing) a dedicated server, such like a Cobalt, so I can build my own projects with the freedom of customizing my own system.

    I think I will go with a Cobalt Raq3i, any recomendations where? Or an alternative?

    Fh

  • by Tim McNerney ( 7987 ) on Monday June 19, 2000 @10:40AM (#991549) Homepage
    The first thing to do is determine whether you are going to need full administrative services or whether you are just looking for bandwidth. Both are available, but make a number of differences in what to look for.

    Once you've decided what you want, find out who else uses a given service. This applies both if you are looking for bandwith only or are looking for full service. You want access to be fast. You want it to be reliable. Do traceroutes from different locations to determine response times to the other clients of the service you are looking at. You can figure out a lot about their peering arangements and see if there are problems with a given site.

    Ask them directly about their peering arangements. Find out about failover strategies. Ask them about service guarantees and make sure that partial refunds on service are available if those marks aren't met. This is key. If they don't meet their obligations, they don't get paid.

    If you are also looking for 24/7 admin, find out how familiar they are with the apps you'll be using. Find out how they monitor the apps in question. Do they have people on site 24/7 or is most of the work done remotely? Ask them about their backup strategy and how they go about recovery of a fully lost system. Ask them to recommend a configuration for your site. You don't have to use it, but it should give you a good idea of how well they understand the networking issues.

    Especially if you are doing the admin yourself, you need to know where the physical facility is and what kind of access you can have to it. If you have a server at an ok prompt, you'll need to get to it physically (unless you've set up a portmaster with remote access).

    That should help you get down to a few possibilities.

  • by adturner ( 6453 ) on Monday June 19, 2000 @10:41AM (#991550) Homepage
    Having had servers at 3 totally different types of Co-Lo's, I can say that you're absolutely right, they are all basically the same. All claim to have the "best network" and uptimes, but reality is that problems will happen no matter where you go. Routers and switches will and do blow up, peering points will and do get saturated, fiber lines will and do get cut. UPS/disel generators will and do fail to keep your servers powered.

    Where things are different is the type and level of service you'll get. Some co-lo's are just that- straight co-location. Others provide more managed services. Some claim to have really good managed services, but really really suck at it. Others specialize at managed services and suck at plain co-lo.

    Another thing to worry about is the fish/pond issue. Are you a big or small fish? What sized pond do you want to swim in? Small fishes in big ponds tend to get the shaft, but their co-lo's tend to have more negotiating power with other carriers. A small fish in a small pond will get better service, but their co-lo may not have the power to get better connectivity when they need it.

    Your most important thing is references from other companies with the same needs as your own. Talk to them at length and get the real scoop. When something broke, how quickly did it get fixed? Did it happen more than once? If they were down for a period of time, did the co-lo's SLA cover their lost revenues/good-will with their customers? How well were they kept informed with regards to the situation and scheduled maintenance.

    My .02:
    Exodus- good for straight co-lo. Lowsy managed services. Has a great or horrible backbone depending on who you ask.
    NaviSite- good for high-end managed services. Lowsy at straight co-lo. Very different network design which may be a good or bad thing depending on your point of view.
    AboveNet- Seems to be a nice pond to swim in, but they seem to have more than their fair share of BGP issues than most.
    GlobalCenter/Frontier- Seems to have a lot of peering problems. No idea on their level of service.

  • by machpo ( 163492 ) on Monday June 19, 2000 @10:41AM (#991551)
    they *all* suck. the best you can do is try to wrest a decent SLA from them, so that they pay for their mistakes - it's small comfort when your pager goes off at 3am because their power glitch hosed a server. They have the latest, greatest UPSs? Great! But you'd better install UPSs in your cage, too. One prominent colo dropped power to our cage 3 times in 3 months. Don't expect to pack your cage to the gills. Many providers have inadequate cooling. It got so bad at my last job that i bought a large LCD thermometer, and put it in our cage facing the main doorway. I had a silent hope that at least one prospective customer would see a reading of 82 deg. F, and ask some questions. You can shift the ventilated floor tiles around all you want, but unless you have adequate cooling power in the first place, it doesn't make a difference! (hint,hint) Don't rely on their expert technical staff. The big colo players seem to have the lowest-paid drones around. To do anything much beyond power-cycling a machine is risking more problems than you started with. And in many cases, that goes for their "security" staffs, too. enjoy!
  • Is it common for a coloc service to require your root password if you are having them host a *nix box? A friend of a friend's company is insisting on having root in case they notice any suspecious activity on their network. I guess I can see both sides of the coin here. I can see how they would want to be sure their network is secure, but I also want to be sure my machine cannot be messed with by anyone. What do you think?
  • by gempabumi ( 181507 ) on Monday June 19, 2000 @10:44AM (#991553) Homepage
    Well, if you're colocating, meaning you are administering your own server, I would focus on one thing: Connectivity.

    Granted, 100% uptime is an impossible goal. But, assuming that you can keep your server up and running, you are dependent on your colocation service for keeping your server connected to the net. The main factor here is connectivity. Make sure they are connected to multiple backbones on different carriers. Look at the network maps of their carriers and make sure the physical location of their colocation service is located near those backbones. Don't get into a situation where you depend on a single trunk line. Try testing their connectivity with a service like netmechanic.com as well. The numbers you get from the test may not be that accurate, but they'll let you know if there's a problem.

    Other issues: bulding security, rack security, multiple power feeds and backup generators, dry fire suppression system. Ask them what their disaster plan is. And, of course, talk to some of their existing clients to get some feedback. Also, if you get the feeling they offer "sales support" rather than real support, forget about it. (a good sign of this is that a call to the sales number always gets through but a call to the support number more often gets a machine.)

  • It would be better if you provide a link to your site.

  • We use multiple facilitys in our work. It depends what you want. Do you want someone else to manage your hardware and software? We do our own maintenance and have found that we get the best connectivity from Above.net and Globalcenter. Exodus is ok but they want to manage your stuff for you. So far for pure colocation I have had the best experiance with Above.net. This is not to say that any facility has been up 100% of the time but the few outages above.net has had have been the smallest ive encountered. There noc monkeys can flip a power switch and plug in a keyboard and type a few commands and even swap tapes but thats about as far as they go on service. They use open alluminum racks for almost all there customers. Larger sites can buy cages (i think 4 racks is the smallest). They have good geographic dispersion and will work with you to multihome if your looking at it. Globalcenter will also do this and uses the same alluminum racks. Both have good physical security as well. Bandwidth in either facility is good however it seems globalcenter oversubscribes there generators which is bad if the power goes out for a long period.
  • 1. good customer support. I don't want someone shouting at me because I called 3 am in the morning. Deal with it, the world doesn't stop. I'd prefer a organization which admits errors and try to fix it rather than those who claim nothing is wrong and put a "all-is-well" smile when the network stinks. :)
    2. less downtime. Anything less than 24hrs downtime in a day is good :) but I'll prefer the ones which stay up longer than that. Ultimately it boils down to how much revenue u are loosing becuase of the downtime.
    3. accessable on phone/physically. If you can't call them up for a mid-night emergency reboot, then whats the point ?
    4. secure cages/racks. I'll not go to an ISP which doesn't ask for my ID when I enter(atleast for the first few times). And I hope all the cages/racks don't have the same key :) (yes its happened before)
    5. big backbone. Just incase... prepare for the slashdot effect.
    6. multiple backbone connectivity. prepare for the DDoS too :)
    7. secondary DNS support. Prepare for HDD crashes...
    8. multiple installation sites... for distributed hosting. If you really become big... distribute it among all over the country or world.

    9. I forgot... make sure you don't go bankrupt :))

  • I have a box coloc'd at Interland. It has been pretty good, but once upon a time, had a real multi-day problem. The "free" techs will only reboot the computer for you...nothing else. Seems fair enough, but I suggest you find a coloc that will do just a bit more for your for "free". Here's an anecdote of my experience there:

    Seems someone had been monkeying around my rack, or it had a bad patch cable or something, because my server was offline. After 5 reboot requests, and confirmation that reboot occurred, still didn't come back up.

    I asked them if they had checked the obvious stuff, loose cables, power connection, etc. and then what happened was this: they said it was my software and I had to hire them (or visit myself, highly impractical) to figure out how I misconfigured the box. It was pretty much the same thing I run on all the other machines I maintain - RedHat 6.1 with current patches installed, so I was pretty confident the problem was not on my end.

    Anyway, I got referred to sales to hire a tech. After peeing on the salespersons' desk for a while, we agreed that if it was their thing, they would fix it, and if it was my thing, I would pay them. Turns out it was their thing, they didn't explain the nature of the problem, just apologized for the 48+ hours down time. My "main" client located on the box wasn't very happy about that.

  • as in, Taco, who are you co-locating with? Actually this was covered in a previous story [slashdot.org]. Slashdot moved to Exodus [exodus.net] from Digital Nation [digitalnation.net] (I think). At the time didn't they cite better support as well as being closer to home?

    Seriously, the claims are going to be very similar from co-lo companies. Previous posts [slashdot.org] really covered the technical things you want to look for. But I think the real value is in first hand experience. For my $.02 I have always liked Pair Networks [pair.com], but I don't know if they offer co-location with your own servers any more.

    Ultimately, I think it would come down to:

    1. Support. This includes the overall competence of the company.
    2. Location, so you can physically administer your servers when needed.
    3. Good routing and bandwidth capabilities.

    Or maybe reverse the order of these. Like I said I have always gotten a good feeling from Pair.

  • Posted by 11223:

    The feature you most want to look for in a server is support for NetBSD. Everybody knows that the BSD TCP/IP stack, as featured in Hotmail, Yahoo, and Netcraft, provides the best performance for a site such as an auction site. Some will claim that NT is better, and others will claim that Linux is trendier (and thus better). Don't believe them! We all know that a NetBSD computer on a 10 Megabit pipe is faster than a Linux computer on a 100 Megabit pipe when it comes to heavy, slashdot-quality loads. NetBSD is just a more mature OS.
  • Yeah,
    I work at a dedicated server hosting company called Rackspace.com [rackspace.com] (shameless plug), and we have a whole frieking wall of Cobalt Raq's (1's,2's and 3's). My advice is that you get a Raq if you want to resell web space. If not, a dedicated Linux or FreeBSD box would be better. Especially if you want to customize your system. Cobalt voids warranties if you even recompile the kernel... and you can't upgrade the CPU's on them. With a Linux box, you can go to SMP and get a RAID-5 w/ a GB of RAM if you want.

  • Sf Level3 is even worse,they oversold the new co-lo facility and tried to take back what they had promised .It took several calls to their CEO to make them live up to the end of the deal. The worst part is that level3 is the only major co-lo in SF unless you want to go to an ISP like Verio (even worse)
  • This wasnt going to be a advertisement, but check out this [bel.bc.ca] for more info including contact information.

  • I have dealt with damn near EVERY hosting provider and co-lo provider on the market, Exodus / UUNet / Genutity / PSI-Net / Abovenet just to name a few and i must say w/o a doubt they BEST overall value HAS To be GLOBIX. The globix corporation has an a #1 rated datacenter from zdnet and for good reason. N+2 Redundancy. not too many datacenters have backups for the backups for the backups. Yes thats right Triple redundancy. and their professional service people actually have a clue. Ive been using them for 4 months and i cant find a bad thing about them. Even the people on graveyard shift have a clue. Ask them to go over and reboot your box? no problem they dont just go over and push a button.. they will sit and watch the whole boot, file check drives if need be. etc.. they arent EXODUS!.. Exodus Sucks.... this is just my .02c
  • It's definately a different experience. I have a linux box a few states away that I use for mail. That one works out well because I know and trust the people housing it. My company's large production servers are all co-lo'd as well, and, generally speaking, it sucks.

    If you have someone in your company who REALLY knows what they're doing, it's best to just get root on the remote box and tell the hosting company not to touch it. In every industry job I've had, the people who will most often be near your machines are the ones you least want playing with it. It's definately worth it to work out ways of doing everything remotely.

    Some of the problems we have run into (not related to the hosting company) are things like Oracle installs, which behave differently from staging areas than from CD, as well as bum tape drives, etc, etc. Backups in general just plain suck. We actually back up locally over a dedicated
    T-1, as well as using a 20-tape juke on the remote end for non-critical backups.

    The most fun challenge of all has proven to be printing. WAN printing is a nightmare.

    All in all, for an equivalent installation, it's a much, much larger pain in the ass. I honestly don't know if it's worth it, especially for what larger companies charge to host.

    --
    blue
  • There are several things that you should check -fire supression and detection : how it is set up is it both under floor and above do they have mutiple methods ie... dry chemical, compressed air, then water. Is it a controlled system so that if it is necessary for it to go into action that it will only effect 3 X 3 foot sections. - Physical security : are any data cables accessable by others above or below floor how are the cabnents set up can others see into or get into the closed cabnet. do they have 24 hr security in the form of people and biometrics(hand scanners and the like) -what is the max Amp load you can pull on a single rack... most co-lo's don't support more than 15 or 20 so if you were running 42 1u servers you would pay an ass load more for the extra circuits - how is there redundant power and cooling handled and how many days can they run with out outside support( power fule etc) - how many redundant generators do they have and how do they cycle in to provide power - how many different providers do they have comming into the facility and can you bring you own to provide backup bandwidth -do they have dark fibre available for increased usage or problems with there own lines. - do they provide POTS jacks in the rack and a phone within 6-8 ft of each rack and internet connections (other than dial up) for your engineers when they are on site. And so much more.... Email me if you have further questions
  • No one gets my root password. If a colo facility asked for it I would be out the door with my money before they got done explaining why. If they need to secure their network they need to secure their boxes.. your box being insecure cannot do anything to their network except overload it and if that happens they can pull your network cable out :)
  • 1. Multiple fiber carriers in to the building in case you need to drop a circuit or two down the way. 2. Locked cage or suite depending on your size requirement. 3. 24/7 NOC techs who can act as remote hands, open tickets, let vendor techs in, etc. 4. Conditioned power, UPS, generator, etc. 5. Multiple peers, UUnet, Level3, and GTE/Genuity should all be peers at min. You should be able to get a partial list of peers. You will not be able to see the size of the pipe to each peer. This is usually confidential. 6. Switched port if you are not going to run your own routers. Make sure they can route for you if you want to bring in dedcated lines or that you can add your own routers in later. 7. Get a tour. Make sure there are no hot spots. Avoid anything with walmart fan blowing air to relieve hot spots. It happens. Always check provisioning times. How fast can they bring another 110v 20A circuit? 220 20A? More bandwidth? Larger space? Another cabinet? Some colo's are running out of space, power, etc. Make sure they have a plan to cover growth ove rthe next year or the length of your contract. Kashani
  • If your demands are not that serious, I would look into a local ISP.

    The main benefit is that you can personally go over there and kick there ass if something goes wrong. Also, the server will be right there, and you can check out their environment.


    "...we are moving toward a Web-centric stage and our dear PC will be one of
  • -Bandwidth (and lots of it)

    -Security (A server with important information located off-site can be exactly the kind of thing that keeps you up at night. Make sure the ISP has good security, and develop your own security procedures and systems as well.)

    -Access (Make sure that you can visit your server(s) whenever you choose, not just during some restricted "visiting hours").

    -Reliability (Do they have good power? Do you have a big enough UPS there? Do they have more than one connection to the net? How often are they "down" from the net? Are their routers high quality? Even 1% packet loss can be annoying (and bad for your business or interest) as hell.

    -Service (If you call them up and tell them to reboot your server will they do it? Will they call you if your server goes down? If they have a problem with their internal network will they be responsive in fixing it quickly?)

    You really need to get a good hold on all of your concerns, needs, as well as an understanding of important eventualities (i.e. how do you detect the server going down? what is the plan when that happens? etc.). Once you have all that down, organize it, group it, prioritize it, and then pick the ISP that gives you what you need.
  • My girlfriend works for AtomFilms.com [atomfilms.com] in Seattle. They use Level3 and the colo building is just 1 mile from my office. When I traceroute to AtomFilms.com, Level3 routes my packets from Seattle to San Jose back to Seattle. Huh?? Why?


  • GlobalCenter/Frontier suck.

    The CEO left and the company had been without a leader and it shows. The merger has things bass-ackwards. I usually end up telling THEM that their is a problem with their backbone. WTF is that?

    Unfortunately, All the major players suck at something. The key is to get SLA's. IF you have and SLA that guarantee's a certain response time and don't get it, you can go to the SLA.

    jas
  • Well what peeves me about Exodus is their pricing strategy. They price in 2 dimensions so you can stack your machines up to the ceiling and not pay any more than if you spread them out a little. This stacking them to the ceiling makes it really a nightmare to administer. I think they should price it per processor or something.
  • Internap claims they have "magic" routing software to minimize packet latency, but I've talked to some IT people that scoff at Internap latency claims.


  • I'm shopping for colocation as well (in this case for a financial application), and I'm having trouble sifting through the marketingspeak to figure out who's really got their sh*t together.

    Does anyone know if one of the trade rags does surveys on/ratings of colocation services? I haven't been able to find anything.
  • We have a number of servers colocated with Level 3 in Los Angeles, and AboveNet in their original San Jose colocation facility. We push _A LOT_ of bandwidth, and have seen some of the good and bad of both providers as an exponentially growing Internet company.

    Level 3 has good connectivity. They have a NOC that is relatively helpful, nothing spectacular but nothing horrible. The connectivity isn't "excellent" but it works almost all of the time (I don't have % uptime availability, but this is the Internet. That statistic doesn't work because you might be able to get to some places and not others). But, trying to get circuits cross connected and new power drops takes forever. And it takes a lot of pestering to get people to move on these non-critical but important issues (You're service might not be down now, but you have to be able to expand!)

    AboveNet has had their fair share of problems. We have been immune to their routing issues because we plug directly into 2 of their routers and do BGP with them and not through a swtich. They will only do this for high bandwidth customers, however. This is why AOL wasn't affected when AboveNet had their switch issues. We weren't either. But everyone in their shared colocation was down for effectively a day. The AboveNet service is EXCELLENT. They will do cross connects in a matter of minutes after you submit the formal request, and the same goes for power and other things. They have decent remote hands service, they are willing to work with you. I don't understand why they have so many routing issues, the network looks great on paper and the people they have working for them are very capable. Maybe they are just trying to be too proactive with problems that they are creating new ones.

    Each colocation facility varies. Take a tour of the three or four that you are interested in, and pay close attention to the facility, to the engineers who will be there to help you when you are a customer, and do some research on the Internet. Check out the connectivity with some looking glasses, see if they have honest Network Status pages up on the Internet (this is generally the sign of a very good engineering team that has power to get things done, because they got over the Marketeting fear of exposing outages).
  • I'll not go to an ISP which doesn't ask for my ID when I enter(atleast for the first few times)

    i'd make sure they ask for your ID every time. and make sure that they have some way of verifying you or whoever is still working for the company. no use having a server there and 24/7 access if that sleazy other tech gets in the day he's fired, says he's replacing the server and never returns..."but he showed id..."
    "Leave the gun, take the canoli."
  • FairMarket.com is a B2B auction hosting service that, as far as your customers are concerned, looks as if you are the one hosting the auction(s). I have no idea what they cost, but several large companies are using them to host their auctions. And not all of the auctions are B2B, quite a lot are B2C. You can check them out at www.fairmarket.com.
  • I agree about Exodus... however I am one who thinks they have a decent backbone. Exodus at least leaves you alone and makes it easy to take machines in an out. Places like QWEST require upwards of 4 forms to be filled out to bring *in* a piece of hardware to put it in your space.

    QWEST sucks. Without a doubt the worst co-lo I've ever had the displeasure of using. A call to their NOC gets routed from NJ to wherever your stuff is. This usually takes upwards of 20 minutes... even for something simple. Once the tech get it... pray they can tell the difference between a tape drive and a computer or you are completely fucked. To say they hire some of the most ignorant tech people ever is quite the understatement. They often have faulty equipment, that they don't detect as being faulty because they only monitor from inside their own network, not from external locations. Wow. They are just bad. I recommend against using them... despite their amazing backbone and facilities. The staff and their NOCs are just horrendous.

    ---
    Openstep/NeXTSTEP/Solaris/FreeBSD/Linux/ultrix/OSF /...
  • You should also make sure that your ISP hosts absolutely NOTHING that might ever get picked up by slashdot... Otherwise one of your rack neighbors might end up eating all of your bandwidth.

    Honestly, though, remember this: The more you pay, the better service you'll get. If you pay $100/month for 10 megabit accss to the internet, you'll end up sharing your line with probably 30 other servers. If you're paying $500/month, it might just be 5 or 6 servers you're sharing with. Of course, you'll still need to actually ask and verify that that's the case... It'd be even better if the ISP you chose was nearby yourself you you could check out their facilities occassionally.
  • AboveNet has an excellent anti-spam reputation. Exodus has a very poor one (their abuse director was known as "Keman the Klueless" when he was at Netcom). This can have an affect on you in at least 2 ways: 1. Guilt by association. My opinion of /. went down when they signed with a colo that is so reluctant to take action against spammers. Your customers may be more broad-minded. 2. Connectivity. A blatantly pro-spam colo will probably eventually get a stay in the RBL (Realtime Blackhole List). That could be bad for your business. I've read newsgroup posts from at least 2 companies who claim to have made their vendor decision based on spam policy.
  • Anyway, they only had 28 comments.

    I expect more from /.

    Fh

  • The main benefit is that you can personally go over there and kick there ass if something goes wrong.

    Not only that, but you could bring along all of the "IF I EVER MEET YOU..." trolls to help you... =)

  • Asking 'how many hops to the backbone' is a sure sign of a total misunderstanding of how the Internet network works. It will mark you as being a total rube, ripe for fleecing. I could say "ZERO hops, the backbone runs right through the co-lo space" and _not_ be lying my pants off.

    It would be smarter to simply insist on a service contract up-front which states the performance you expect from them,(X amount of bandwidth available to my server, X latency measured between Z and Y, and oh, I need a lock on my cage!) and how it will be measured (TTCP every other day, a basic ping, and only you and I know the combination!) and how many free months/day/weeks of service you get (or $$$) when they don't meet those performance numbers.(Your main co-lo link was down for 6 hours!!! That means I get a credit of $1200!)

    Think about how small companies get telco's to do work for em? HIT EM IN THE POCKET BOOK when they fail/are late delivering service!

    Depending on the depth of your pocket, Co-Lo ranges from an empty rack with a power cable and no network, to your own private room, with a lock on the door and multiple Gig-E feeds.

  • My friends and I have our baby p75 (soul.apk.net [apk.net]) on a co-located network. What I find is VERY important is a redundant internet connection to several different backbones. I personally think that most networks suck, but when you have a connection to 5 different ones, speeds to various sites are going to be much better. The rest of the suggestions have been very good.

    Mike Roberto (roberto@soul.apk.net [mailto]) -GAIM: MicroBerto
  • You probably shouldnt give them the root password.
    What you should do is ask them to create an ssh key, where they can use the ssh key to login as root, that way you can atleast tell, provided your logging is set up correctly, when, where, and how they login as 'root'

    love

    christopher
  • That site is a START. Let me just say, though, that any web host ranking site that lists Superb Internet as among the top 25 is suspect. Their bandwidth reliability may have improved over the past little while, but only because they have no customers on their lines. Don't plan on having any of their staff available at 4am to reboot a server, and don't plan on ever visiting the NOC or replies to your questions via email or phone.
  • A front door? </joke>

    ---
  • A company I contract at currently is co-lo'd at Exodus (I won't say what site, as I don't feel this is a fair impression of Exodus as a whole - I've had several good experiences elsewhere, and I don't want to taint anyone's impression of them). One of the other SysAdmins had to run down to do something or other to one of the machines at the co-lo and when he walked into the server room, he found the whole rack wrapped in a plastic tarp.

    Apparently, they had been doing a bit of electrical work near the rack and didn't want to have any debris get in the machines. Well, apparently they either work really slowly or just forgot about it (how you can forget a big blue tarp is beyond me) as the machines were up to around 140 degrees (F) internally (90s externally).

    Ouch

    Yeah, so, onething you should look for is that not happening.

  • InterNAP made no such claims to us - you're the first I've heard to say so. Happy customer here, just curious what the hell you're talking about! :)
  • There are a lot of seperate price factors besides up-front and bandwidth charges. What does it cost to get remote hands to do work in your cage? Do you have to pay for power? How do the bandwidth charges scale when you need to expand? What about how the bandwidth charages are calculated... are they based on average traffic, do they drop the 3 minutes of highest burst, etc. Speaking of which, what about burstable pricing, etc.
  • by Felipe Hoffa ( 141801 ) on Monday June 19, 2000 @11:25AM (#991591) Homepage Journal

    Who Invented It?

    It is commonly believed that the Internet came into existence as part of the United State's government's nuclear war fighting strategy. Originally conceived as a robust and fast communications network called the ARPAnet (Advanced Research Project Association Network), it was designed to help scientists and technology researchers communicate, in fact, it was created in direct response to the Soviet Unions launch of Sputnik, the first man made satellite. It was part of a whole series of initiatives taken by the US government to enhance science and technology development. Later, the multi-routed and redundant telecommunications lines, switches, and computers were discovered by the military, to be an ideal network to prevent Command Control 'de-capitation,' in the event of a Soviet 'first strike' against the continental US. source [hostfind.com]

  • Umm, excuse me, lad. Why is "how many hops the backbone" not a valid question? It certainly is. Unless the company is one of the key internet backbone providers, like Digex, MCI WorldCom, MFS, Sprint, UUNET, WinStar GoodNet, etc, it's a very IMPORTANT question. Because while some other companies might have staunch bandwidth, a good backbone themselves - they might not be as highly rated as the KEY "backbone" providers. Maybe you should read my Bio before you tell me I'm a stupid "rube" person, motormouth.

  • Everyone else has added the important ones, but I will throw in a few that I didn't see:

    • Fault-Tolerent or Highly-Available clusters
    • Power backups
    • RAID backups or equiv.
    • Bandwidth to support expected TOTAL load
    • Intrusion Detection policies & countermeasures
    • Maximum Restore Time, in event of system failure
    • Average up-time, per box & overall
    • Adequate Authentication Procedures
    • Systems using OS' with working TCP/IP stacks
    • Multiple Physical Locations
    • Adequate Physical Security of machines
    • Round-time for operations (incl. manual)
  • Actually - the service that you get will depend on your ability to manage the relationship. All co-lo's will have issues. All co-lo's have poorly trained 24/7 NOC staff.

    You MUST be the one to shove your needs down their throat and get them to respond.

    Having also had systems hosted at various co-lo's I so far believe Qwest to be the best. (they still have their share of problems) but for the most part their facilities are the best - they have a solid, fast backbone and you can't beat their prices.

    Managing your vendors is the key to your success - don't accept what they tell you about outages, redundancy etc... problems happen, and when they do - you need to stay on top of them to make sure that you get all the info out of them as to ensure that you will be comfortable with the resolution.

    Here is a list of the requirements we put together for a co-lo:

    Fast reliable bandwidth.

    Bandwidth provided by the collocation partner must be fast enough to handle the current and future requirements of hosted services. In addition, reliable speeds must be provided through such means as an SLA to ensure that the quality of service provided meets the expectations set by the signing of a contract for service.

    Redundancy.

    Collocation partner must provide redundant connections to the Internet via multiple links and/or multiple ISPs.
    Redundancy must be provided for the following standard services; Bandwidth, Power and cooling. All services provided by the collocation partner must not be a point of failure as to incur downtime or excessive risk to the operations of systems collocated therein.

    Physical Security.

    Physical access to systems collocated should be limited and monitored by both human and electronic means. All access to systems must be documented via videotape or sign-in at time of entry. Cage/Rack access should be limited to an approved access list and ideally accompanied by Collocation NOC staff.

    Network Security.

    Network services must be monitored on a continual basis as to ensure the integrity of the network security/health of critical nodes not within the direct responsibility of Quicknet Network Technicians. All routers, switches etc. shall have measures taken as to prevent, as much as possible, the attempts of denial of services.

    Fail-over Capabilities.

    In addition to system and network redundancy - services must be provided with the ability to fail-over to a healthy system in the event that the primary system is below acceptable functioning levels. This includes such services as HTTP, DNS, DB etc. continued availability of services with minimal impact to the users.

    These are rather generic and you will find that most co-lo services will say that they can meet these - but remember the two most important aspects of hosted services:

    1. it is a sellers market. They can and will get whatever they want as there are so many customers and not enough hosts.

    2. the fact that it is a sellers market doesn't mean that you are at a disadvantage. This is a new industry - and since most of the companies are inexperienced, you can pretty much manage the services you get by staying on top of the implementation manager and the account manager. They typically dont have too much experience in dealing with hosted service, they do want to do a good job and if you clearly communicate and document all requirements and expectations you will get the best results.

    Exodus: HORRID! had NOC staff walk by and trip on the power cords to machines - taking me off line. No secure space - unless you get a cage - all racks are open to other customers in your area. BAD. Dont own their own backbone (which means you pay more) max bandwidth oc 48. worst managers in the industry. (tech staff forgot wr mem on routers - then when they reloaded - they had lost the configs and many other horror stories....)

    Frontier Global: ok - cant say too much bout their NOC staff... own some backbone. but account managers are not too responsive. oc 48+

    Intel: Fully managed, they only want to be the full asp and dont yet have the option for strictly co-lo. heard they had an oc192 - but not too sure...

    Qwest: great owned backbone. ok NOC staff, great account manager/implementation manager. Some internal system issues with their extreme networks switches. but due to their close relationship with extreme - they tend to get solved fast. Also - extreme has a deal with F5 to put F5 load balancing on an asic in their switching routers. this means that their will be no need to buy your F5 in the future - just pay a little more for the switch port. max bandwidth OC192. plus they have a lot of dark fibre - so they wont get saturated any time soon. SLA says 75ms anywhere on their network. Fully locked enclosures, easy to get what you want.

    my $6.73.

    Drinking is the act of God pouring God into God.
  • by HavokDevNull ( 99801 ) <eric&linuxsystems,net> on Monday June 19, 2000 @11:29AM (#991595) Homepage Journal
    I work at a small Co-Lo and web design company in the North West. And have worked at large ones as well. And from my experience it is not all about who has the biggest and "fattest" pipes or bandwidth. It is about the people.

    Ask yourself this question would you have your server housed with 100 Network Operation people running around with their heads cut off, do not know your name, and would not give you the time of day unless you flashed $100 in front of their noses, because the are so big.

    Or a small company, were you know they will know your name, and are able to take the time to hear your concerns and not pass them up the food chain to middle management. I personally know every client by name and their box even better. I know each of the boxes personal quirks and their needs as well as the box owner's quirks and needs. Because it is a small company I'm in better position to do the job I love and able to pass that along to the customer.

    So my suggestion is to meet the network operation personnel the real people taking care of your investment, not just the sales manager. Ask to take a look at the server room. Make sure it is clean and free from obstruction, environmentally controlled. Do not let some yahoo tell you that you need dual oc-48's to host your site. I can host a site just as well as anybody from my dsl-line. Get a feel on how well they take pride in their job, and you should be fine.

    Which one is better?
    "Kill -9 needs no justification" BOFH
  • After you've decided who to colo with, when you sign the contract, make sure there are provisions in it for you to make X number of spotchecks of their services per month... and make sure you're able to do it without telling them ahead of time. ;) Remember, if they say they're up 99% of the time, keep checking on that figure! Remember those fire-drills you used to do at elementary school? Apply the same technique to your colo.

    Chris
  • One key concept people overlook: when the site goes down at 3 AM how bad will the drive over suck? When the site goes down right in the middle of rush hour, how long will it take to fight your way there? This WILL happen, and you need to make sure your hands-on people have good access.

    The other thing to try out is doing some traceroutes to sites you know are already hosted in the facility. Some of these places are a REAL MESS due to the speed they had to grow the internal network to keep up with business. If you see 7 hops from their front door to the customer router, be very scared.

    Another trick is to check out how professionally done the customer cages look. Slick wiring jobs? Beautiful rack mount setups? If the other IT monkeys at the place look professional, they problably didn't pick a looser to host their site.

    As far as site security, power backup, etc all of these places will be selling you the same thing.

    Good luck!
  • According to their webpage [csoft.net], csoft.net offers colocation for $30/mo. I am not running something extremely mission critical, but high uptime is nice. Does anyone know if this is for real, and if it's any good? They don't seem to respond to emails.
  • The following is a little paper I wrote on the seattle colo market last Apr.

    Before we get stated let me just say that I HATE internet colo centers (ie exodus). I really perfer the telco colo areas. But they arent caged (one of the reasons I like them) and normaly use open telco racks. I have been in more than a few of them. I would do anything to get my employers data center out of the noc monkey ran exodus. Shit we lost power a week ago and they didnt even contact us because 'we wernt one of the effected customers'. Well we were and they had our power labeled wrong. Not to mention that the monkeys behind the glass never want to get off thier ass and let me in.

    If you want high quality bandwidth and care about colo second find an internap.com pnap near you. Trust me that it is good. I used to build there pnaps. Not to mention 12+ backbone connections at each PNAP. The PNAPS are works of art and probably better cabled and labeled than your stuff (isnt it soposto be that way?) Have the sales guy show you a pnap. The only thing I ever saw that came close was the computers at NORAD (cept norad's computers were from the late 60's [can you say iron-farite core memory]). Hey looks arnt everthing but you when every wire has a pre ordaned path you know someone cares.

    ---
    colo.txt
    Co-Location space in the Seattle area is currently in high demand.
    Customers are currently filling it as quick as providers build it. This
    is not only a Northwest problem. Switch & Data Facilities Co told me that
    in the last 2-3 month they have gone from virtually no customers to
    filling out there 11,000 sq. ft within the next 4 weeks (they are approx.
    70% full right now).

    Neutral Co Co-location sites:
    SDFC.net
    Titan
    Colo.com
    Nextlink
    Level3

    Internet and co-loaction:
    InterNAP
    Exodus
    Savvis

    Switch & Data Facilities Co
    Bill Barnard of SDFC was one of the more useful reps I talked to.
    He not only knew his prices (and was willing to do over the phone quotes)
    but he knew the prices for most of his competitors and was not worried
    about quoting them to me even though most everybody else was less. SDFC
    is the only company currently operating with there business model. They
    offer truly neutral co-location space. Because of the telco space problem
    around the country right now many of the big us and foreign carriers
    (including Deutsche Telekom) are putting there equipment at SDFC. SDFC
    space is on the high price because of the demand. They are charging
    around $1300 per rack with a 20A AC feed. The are located in the Westin
    building but have a north end location planed and are looking for a south
    end (Renton - Auburn) location. The Westin location will be full in 4
    weeks (1 May).

    Titan
    Titan is unique in the fact that they use old hardened military
    sites. They are made to be resistant to earthquake, bomb blast, EMP.
    They have other interesting features:

    Electronic perimeter security systems
    High security exterior lighting
    Mandatory escorted accessFire resistant interior walls
    DoD authorized "top secret" vault

    The only problem is that old hardened military sites are not generally in
    the middle of metroplexes. Titan1 is no exception, it is located at the
    Moses Lake airport. So I believe this doesn't fit out need for a 30/2
    response time. But Titan may be a good choice for off-line tape backup
    storage where response time is not important but survivability is.

    COLO.com
    Colo.com is building there new space in the USWest/ATT downtown
    switch (3rd and Spring). There space will be completed in June and than
    will be building sites in Bothell and Tukwilla. One problem with the
    downtown switch is that it is a no cell phone zone. When you enter the
    building you are required to turn off you cellphone (and it wont work even
    if you don't turn them off). I know when we were at the ATT facility up
    on the 13th floor we were told that this wasn't a 24x7 facility. I don't
    know if that has to do with the downstairs security guard of if that was
    just ATT. Colo.com will be selling both rack and cage space.

    Nextlink
    Both Nextlink and Level3 are at the 1000 Denny building.
    Nextlink is almost full and will be full shortly. I was told that 2 racks
    (not cabinets) would not be a problem but cage space or much over 2 racks
    would be a major issue. Note that Internap seattle PNAP2 is located in
    this space also. Racks go for $750 each.

    Level3
    Level3 is basically full. I think mostly they are selling people
    into Nextlink space. Also level3 is not really neutral because I am sure
    that they would want to sell us their internet and phone service.

    Savvis
    I used to work for savvis so I know / built their data center.
    Savvis does not currently do co-location in Seattle but I have heard they
    are thinking about building out the rest of their data center do to it
    because of the space shortage. Savvis is located on the 27 floor of the
    Westin Building. I have had recent problems with Savvis's NOC being
    unresponsive and unknowledgeable.
    Last update
    Tue Apr 11 09:32:20 PDT 2000

  • Colocation.

    Not virtual hosting.

    I'm not.

    With colocation, you're no longer sharing machines with other sites, but you still are sharing the pipes... If one machine gets, say, slashdotted, and your host doesn't have enough bandwidth set aside to handle it, your machine is going to go without it's bandwidth until interest in that machine dies down.

    Whatever. If someone advertises a 10 megabit (ethernet) connection to the backbone for your machine for less than $1200-$1500 a month, you're going to end up sharing that line with other people. You just need to make sure that the place you choose is either big enough to be able to handle the demand, or that most of their machines/sites are small enough that you'll never end up fighting for bandwidth.

    Another example: Avoid colocating at places that have more than a few servers hosting adult sites.
  • > The problem with colocation is that everyone
    > promises the world, but non deliver.

    You can say that again! The problem is bigger than that... there are lots of colos that are "big-time" colo facilities. They all suck. But mostly they suck less than doing it yourself, and some suck more than others.

    > * switched ethernet (ok, if you're talking 100
    > meg colos... but if you find you're on a hub,
    > thats a major sign they are incomentent)

    Worse that that... you end up soaking up somebody else's problems. I talked to a guy just a week ago who spent three weeks diagnosing why one of his customers wasn't getting the throughput he wanted. When he was finally able to get the data center to look into it, they figured out that since the network wasn't switched, some OTHER customer had a bunch of NT boxes doing broadcasts and clogging everything up.

    Make SURE it's switched. Most are, but obviously some are not.

    Basically -- if all you want is hosting, go with someone who speicalizes in that -- like a verio or an Epoch. If you're going to have multiple dedicated machiens of your own in the colo, then you gotta go with a full-service colo.

    Use a service like Service Metrics to investigate the colos response-time from various points around the country so that you know its weaknesses. (I think there's a freeware version of this these days too.) Also talk to other customers to find out what they think.

    Good luck!

    jim

  • While you are at it, try asking, "What can't you offer that others will?" or, "What are any problems for using this service?"

    Actually, these two questions work well at job interviews, potential graduate school visits, and many other places. Use them wisely.
  • Oh, yeah. Like you're going to ask a marketing guy whether his NOC equipment is on the floor. I don't think our marketing guy has ever seen our NOC, so he probably doesn't know that it has fire sprinklers in it. He also probably doesn't know that the building is on a flood plain.

    Actually, you'd probably get the best results if you called tech support, asked them a ridiculously weird question, finally asked for the NOC administrator, and then asked him questions like that. But, then the marketing guys would probably hate you and the NOC administrator, in his perpetual state of caffeine-induced agitation, would be so perturbed that you wasted his time that he'd make your life a living hell.

    In conclusion, the only way to know whether or not to colate somewhere is to put in a job application, work for them for three weeks, and then quit.
  • Well here's a followup shameless plug (hey, he started it..)

    I work for another hosting company called Net Infrastructure [netinfra.com]. I agree with Skratch; Cobalts are great for webhosting, but not much else.. If you want customizable, go with a cheap Linux box deal which you can upgrade later on if need be.

    -zmn

  • No way, no chance, no how. Nobody gets my root access. If they have an issue that is truly causing problems, they can always disconnect you. That is, if they can't reach you beforehand.
  • Some things you can ask a provider for (in addition to the bandwidth and storage considerations) that might help you make the call: 1. A BCP plan, a good one should include some indication as to the likelihood of natural disasters etc. 2. Security policy and escalation plans. 3. Insurance information. How well are they bonded etc. 4. Additionally, it is important for some companies to know that the background check and HR process is at least as stringent as their own. 5. Ask for the results of their latest security audit. Just a few off the top of my head. Good luck.
  • Whelp - read my other post. It all depends on YOUR ability to manage them.

    Here is a hint on how to get around the qwest call center:

    Next time you are in the facility - talk to the NOC staff and the NOC manager. Make sure that they KNOW you. Then take a look at one of the phones in the NOC. (they should all be labled with their real DID #) memorize it.

    Next time you need something - call INTO THE NOC Directly. They will usually say that you need to open a ticket - you reply "I will - but I need this done now. I will follow up and open a ticket as soon as you do x."

    Works for me every time.

    Also - make SURE you keep your account manager and implementation manager informed of every little detail. When you dont get a response - email the hell out of em. and copy all you internal staff (obviously just the ones who whould be included) and then they can reply directly to the qwest ppl.

    Basically - know that the qwest employees dont want to look bad - and they will jump through hoops to make sure they dont look like they are the reason your pissed. This really gets internal qwest moving :)
  • Lots of places do the right thing, with regard to bandwidth, routing, contigencies, etc. Many of them suck, however, when it comes to things like remote support, or simply not screwing you accidentally. You should ask for references, having a profile similar to your company, and see what they have to say about the service. I don't know if I'd trust a babysitter, or a Colocator that wouldn't provide references. Fat pipes and low hops mean less if some idiot named Chuck keeps unplugging your uplink.
  • Colo's a fairly wide business, with different providers offering different topologies, and hosting businesses piggybacking on colo businesses, professional services businesses consulting for both, some colo provided by IP carriers, other by pure colo providers. You need to think about what you're trying to accomplish, what your busines needs are, how much management you want to do, and anything special you want beyond the vanilla service (e.g. lots of electricity, access to local telco facilities, etc.) Here are a few categories of services and providers
    • Shared hosting - you're not renting a box, you're just renting capacity on a box.
    • Dedicated managed host - the provider is still managing the computer, but it's all yours
    • Dedicated unmanaged host - you're renting the machine, but you're root; extra fees for hands-on help. You may or may not have physical access.
    • Cage/Rack rentals - You're renting real estate, power, and network feed. Physical access is usually somewhat restricted, but some places let you do whatever you want in your cage. If you need more power than a vanilla colo, e.g. you've got a lot of 1U servers, check with the vendor carefully - some places can't handle it, or can only handle it in some of their buildings.
    • Content Distribution/Caching Services - This is an alternative to using your own systems to distribute everything - pre-cached or demand-cached servers handle lots of the content, especially static graphics. Akamai is the best known, but other players such as AT&T are getting involved, and everybody's got their niche.
    • ------
    • Pure-Colo/Hosting, Few Locations - some companies aren't in the datacom business. They're generally located in one or more NAP/MAE cities, and rent telecom from big ISPs. Check out their service providers and peering, but think about your performance needs - for some customers, the extra few milliseconds of response are critical; for others it's the quality of technical support, or price or quantity of the raw bandwidth.
    • ISPs providing Colo - Level 3, AT&T, Frontier Globalcenter, etc. - these providers have backbones, and customers on their backbones, and may be providing peering from their backbone rather than their hosting centers - or they may do both.
    • Telco Cage Space - AT&T and some Bell telcos rent cage space in their offices. If your business needs a large number of physical locations, or better connectivity to the telephone network than colo vendors provide (e.g. for modem pools, or DSL, or lots of T1s, or lots of DS3s or OC3s between your locations), these may be an interesting alternative. They tend to be extremely secure - and therefore hard to get access to at times - with exceptionally good power systems, fire/earthquake/flood resistance, and cooling. They're usually more bureaucratic to set up, e.g. needing to know power and HVAC needs upfront, but they're located almost everywhere.

    (Disclaimer: I work for AT&T, but not in the hosting group., and this is my personal commentary, not a company statement, in spite of the occasional shameless plug in the content.)

  • by Tony Shepps ( 333 ) on Monday June 19, 2000 @12:11PM (#991613)
    "Multiple OC3s! Automatic Halon systems! Passkeys and dead-space entryways!" Very geek of everyone to require highly expensive but ultra-cool setups of their providers.

    Did anyone see Jurassic Park? OK, dumb question, everyone saw Jurassic Park.

    All these ultra-complicated solutions sound very Jurassic-Park-ish to me. There is, it seems, a turning point where you have such a complex setup that it is guaranteed to fail; and when it fails, it will fail in unpredictable but spectacular ways.

    Like, as someone else pointed out, the colo that was so secure that he couldn't get in with a valid ID card. Doesn't it defeat the purpose of security if the people who should be able to get in cannot get in? Isn't that just as much of a danger?

    And what's faster -- multiple T3s that are saturated, or single T1s that aren't?

    And exactly how long does it take Cisco's biggest router to boot? And how many of them are between you and the rest of the world? And why would you want that?

    The biggest difference between ALL co-los, in my opinion, is customer service. THAT is the value added that really makes a difference. Your system is unavailable at 3 AM. DOES ANYBODY CARE? This means so much more than all the geeky stats, hops off the 'bone, etc. If they CARE, they will put you right, no matter what their situation; if they DON'T care, you will be unhappy even if they have all the gear in the world.

    So, evaluate your potential providers the old-fashioned way: ask other customers if they are satisfied.
    --

  • by Silas ( 35023 ) on Monday June 19, 2000 @12:14PM (#991615) Homepage
    A related question...what should co-location services reasonably be able to promise?

    I run a web hosting firm that currently does only virtual domains, but is looking at starting to offer co-location services. We're trying to figure out how much headache/cost is associated with each "level" of service: UPS backups, 24/7 support, multiple net connections, sysadminning, etc.

    For those of you who already offer co-location services, what's reasonable to offer your customers who want an "average" (i.e. not ridiculously expensive) co-location experience?

  • by Raptor CK ( 10482 ) on Monday June 19, 2000 @12:20PM (#991617) Journal
    Having seen my fair share of datacenters, I can't give you any reassuring advice.
    1) Exodus (NJ location) looks obscure, but lacks any decent security persence to the extent that I'd like. You just sign in to gain access, and all that they check that on is a contact list.
    2) GlobalCenter (NY) uses the same system, but there, you can sign in as "Samuel L. Jackson," "Tommy Lee Jones," etc., and they let you in anyway. They don't even check.

    These two also share another aspect that I dislike: Communal rack space. You can have your servers right next to those of another company, and there's no barriers at all. Makes it incredibly easy for gear to get inexplicably damaged, although it's rare that it happens.

    GlobalCenter's got a decent amount of private cages available, though, and those are pretty good.

    3) Level 3 (NY) has lockdown cabinets in the datacenter, decent security, and biometric identification for datacenter access. If you don't belong, or aren't escorted by someone who *does* belong, you're not getting in very easily.
    However, they take forever to actually get your line installed sometimes, which certainly puts a damper on your critical business setups.

    4) Globix (NY) has much of the same setup as Level 3. Lockdown cabinets, biometric identification, etc., but they also have beefier physical security, and even if you *do* belong there, you still have to sign a key out if you want access. For some systems, you'd need to authenticate across no less than 4 keycard locks, 3 of which are biometric, two security guards, and a room full of NOC techs. The drawback? Their turnaround time can be dog slow as well, and they have been known to oversell well beyond their frazzled tech staff's capabilities. To make matters worse, the datacenter has windows, and you can see their gear from across the street. That's extremely discomforting.

    As for key questions, I'll bypass most of the redundancy, and just recommend a tour, possibly inviting your own network engineer or a consultant to help out if you don't know all the BS from the actual details. Key things to ask about are:
    - Raised floors / Drop ceilings? (Key for them wiring you quickly and flood avoidance)
    - Conditioned power/Backup batteries/TESTED generators?
    - Glycol or Halon fire suppression? (Glycol is less likely to kill you if you're in the room.) To that end, ask if there's a way to halt the fire suppression if you're still in the room. I can't stress the importance of being able to get your people out before the FM-200 kicks in.
    - Security and surveillance? You need ID checks, sign-ins, the whole nine yards, since nothing is more critical then your information and the gear it runs on.
    - Network redundancy and quality. I'd focus on the redundancy more than the quality myself, since most providers do have pretty much equal bandwidth. However, if they only have one way out of the datacenter, that's going to hurt you at some point.

    Those are the key points that I'd worry about. NOC competence isn't something you can always rely on with any company (I should know, I *was* one :-), so just be sure that you know your stuff well enough to get things done quickly and clearly.


    Raptor
  • It [colocations.org]'s a nice site (the sister site of shellreview.com) that has reviews of various colocation services. The admin really does some nice, thorough reviews. While I wouldn't reccomend it as a primary source of information, it probably would be good to see if the companies you're considering are listed here.. And if so, if there's anything nasty said about them.
  • I haven't tried them for colocation services (although I've discussed it with the webmaster) but I heavily recommend Seagull Networks [seagull.net]. I use them for hosting several domains I own and always recommend them to people who ask me.

    Here's why:

    • They allow shell access via telnet and secure shell
    • Supporting ssh allows me to use secure copy (scp) to upload content
    • I can read my email via a shell login with Pine or Elm without downloading all my mail (important when one uses several operating systems)
    • I can write my own CGI's in any programming language I want and install them myself. They provide the gnu development tools.
    • They have excellent customer service. I've sent in questions in the middle of the night and got back authoritative answers within the hour.
    • Their prices are quite reasonable - $25 a month for basic virtual domain hosting, which might seem high but you get the shell access and secure shell
    Write to seagull@seagull.net [mailto] if you want to ask the webmaster about colocation. That's not really their specialty but hosting is.

    I host these domains with them:

    In addition my fiance has two domains there and a friend has two domains there under my account (there's a discount for reselling the service - your first account is free but more under the same billing are cheaper).

  • by Wicked Panda ( 10814 ) on Monday June 19, 2000 @12:50PM (#991632)
    Then GO THERE! A physical inspection of the facility is always useful. Inspect where your stuff would go and how secure it is. I don't want to think about how many outages we have had from a NOC monkey messing with wires at the patch panel.

    Monitoring is always useful, and also check to see what kind of access you will have (better be 24x7).

    People have already stated about power and cooling, but having been on the wrong end of this before - make sure they are redundant!!!! One generator dying and taking down a datacenter is unexcusable.

    One thing I didn't see mentioned, is when are the maintenance windows for the infrastructure that you are sitting on. If they have to do a router firmware upgrade - they better do it in the middle of the night.

    OK, thats my $.02
  • I call a large amount of BS upon you, AC! There's a lot more to what they do over there than tier one peering. There's also a high-quality NOC out there, as well as some route optimizing stuff that whups mightily upon some of the BB providers own network diagrams. in addition, they recently just purchased some large co-lo company, Co-Space, and so they're only getting better...

    For my 2 cents, not that it matters down here in moderator land, but I think the best way to get a quick fix on the quality of your co-lo is to call and ask them to locate a particular machine. Some providers, believe it or not, are incapable of actually doing this. I've seen a night tech out at verio essentially say "uhh, to find that machine, I'm going to have to just shut down the connection and wait to see which CSU/DSU lights up..."

    Jeez.

  • Also practical to know whether they have surge protectors (on the ethernet/phone lines as well) and/or UPS'es; if so, how long will they keep up the power, and if it ever fails, do they have a separate signalling cable for each computer, and whether in case of shutdown will they restart the boxes after the power situations stabilized.
  • Anyway, notice how WonkoSlice links to Slashdot, and Kuro5hin mentions Slashdot in their FAQ (and probably other places), but Slashdot never makes any mention of Kuro5hin or WonkoSlice or any "Slashling" sites, except in a few user's signatures and in offtopic troll-like posts like this one. I wonder why..

    Cause Slashdot was first to make this a viable form of website. We all owe them that, and any one of the "slashlings" that says they owe nothing to Slashdot is lying. K5 owes most of it's audience to slashdot as well.

    Didja notice, though, that /. has a Kuro5hin slashbox? It's not nearly as confrontational as everyone seems to think.

    --

  • It would be nice if you would move the line

    I work at a colo company, and our fast facility meets these requirements...

    to the TOP of your advertisement, so that we would know right off you are just trying to sell something, not give usefull information.

  • Anyone have experience with DellHost.com? I'm seriously thinking of going with them. Huge brand name, which doesn't mean THAT much, but at least they won't go out of business and the accountants will be happy. Plus, there prices are great: $300 for a 650 mHz PIII with 2 9 GB SCSI drives and 256 MB of RAM, plus 21 GB/month of metered bandwidth to start. Bandwidth upgrades seem quite reasonable too.
    It almost sounds too good to be true though, as even very small ISPs haven't been able to beat those prices. Is this going to turn out to be a horror story?
    Thanks!
    --JRZ
  • by treat ( 84622 )
    Exodus has been mentioned several times.

    Their cages, at least in New Jersey, have only about 5 different keys. That is, every key opens about one in five cages. Try it.

  • full BGP4 routing so all those pipes are used at all times, not just when one fails.

    Here [networkcomputing.com] is an article describing what BGP4 (Border Gateware Protocol) is and why you need it.
  • Use a service like Service Metrics to investigate the colos response-time from various points around the country so that you know its weaknesses. (I think there's a freeware version of this these days too.) Also talk to other customers to find out what they think.

    "Exodus Acquires Service Metrics For $280M" [techweb.com]

    Geez.

  • ... indivdual locked cages ...

    One Co-Lo service I visited had nice individually locked cages, all ordered from the same manufacturer. Except that the manufacturer was cheap, and all the cages had the same key! So much for physical security.

    What was really funny is that I discovered this while walking through their machine room as a potential customer. You shoulda seen the sales rep's eyes pop when I took the key from one cage and opened another one.

  • What are the bandwidth differences between these pipes?


  • by cpeterso ( 19082 ) on Monday June 19, 2000 @02:33PM (#991655) Homepage
    DS3 = 45 Mbps line
    OC3 = 155.52 Mbps line
    OC48 = 2.48832 Gbps line


  • Perhaps one day you'll realize that some people actually do what they love for a living. Forgive me for knowing what I am talking about and I suggest you follow a similar path for greater satisfaction in your life.

    Please feel free to compare our products and services with any that you'll find anywhere else.

    In an industry which is filled with uncertainty it is still relationships with people which drives business. People like myself offer advice and information to the best of our knowledge, freely and in the hope that it will benefit the electronic community.

    -ryan
    I might add that the posting I wrote, the last sentence was last because it was an afterthought.

  • While I don't have a specific answer to the question, I do have some thoughts on things I've learned along the way.

    I once had a friend at a start-up ask me whether they should colo or do it themselves. I have an edited reply here: http://www.nspf.net/colo.txt [nspf.net]. In it, I talk about several things that they or a colo provider would have to think about and plan for if they were building and managing a data center. As you look at a data center, you might want to think of some of these issues as well.

    One of the older reasons that people started using colocation was that local loop charges from an office into the ISPs was expensive. By putting a server at an ISP, you avoided those charges. The data centers at the colo were better than one's typical office. Once you get past DS3 speeds, the LEC charges don't look as expensive as they used to. Smaller sites (<=10Mbps) are still better served at a colo facility.

    Recently, the reasons for picking a colo provider have morphed into: "What am I willing to outsource?" When you pick a colo site, you're trusting your machines, networking, facilities management and physical security to another company. Are they better at doing it than you?

    Some colo providers distinguish themselves in various ways. One might be better-connected into AOL or broadband networks. Some focus on connectivity. Some might be better aquainted with NT than others. Some might have a great daytime Sun or Linux staff. One might offer database outsourcing/management. One might offer backups. One might rent you EMC/SRDF disk space. One might offer managed servers instead of using your servers. These features could good reasons to select one over another, but only if they complement your operation. Make sure, though, that you don't get attached too much to any colo provider's value-adds. For example, if they have network problems every week and your data is trapped in their managed database, you'll have to live with it until you can duplicate the functionality that you'll have to leave behind.

    Initially, in an R&D and trial/rollout phase, many companies can get away with outsourcing as much as they can to focus on their site's development, but as the site becomes more popular and as users depend on it more, the uptime and reliability of the site becomes much more important. You can't just let your colo or network provider screw up anymore. It's unacceptable. Decrease your dependence on any colo provider as your site becomes more important.

    A very annoying feature of running a colocation site is working remotely. If you need to do anything more than hit a power button on a machine, you need to figure out a way to not have to do that. Don't change tapes, buy an autoloader with lots of tape capacity. Don't use machines that need keyboards/monitors - use serial console servers to access the serial consoles of your devices or use something like Citrix to manage your NT boxes. Buy reliable machines that don't crash as often. Inside a server, a 10000 RPM or 15000 RPM disk tends to fail more than a 7200 RPM disk. A beefier power supply running at half capacity runs longer than a cheaper one running at capacity. Buy more servers than you need for everything so that you can migrate your service from failed servers to standby ones. Don't run the latest version of an operating system. Run the most stable/patched version of the OS. Eliminate all single points of failure from the networking side (including having more than one upstream ISP if possible). Routers and layer4 gear need to reboot sometimes. Always buy more than one of each.

    The best way to avoid failure is to have multiple data centers. You can care less about the reliability/availability of one data center because you can always direct traffic to another one. Many web sites make a mistake early on of building a single dependency into their site, whether it's a database or a filesystem, something keeps the site from running in parallel with a similarly-configured. Plan from the start by running your web site from 3 locations, and you'll be able to scale your site very well. You'll also be able to pick and choose from cheaper colo providers that don't do N+1 redundancy to help reduce your costs.

    At some point, every colo provider will let you down. It's inevitable that something bad will happen. Picking a colo provider that learns from its mistakes can be better than one that strives to make none. We would expect the same of our employees if we ran a data center ourselves.

    If you're good at managing servers remotely, the location of your site(s) becomes less important:

    • I have a friend who colocated his servers in San Diego while his team worked in the Bay Area. One of his justifications was that it was just as fast to drive from his office in SF to Oakland and fly a plane to San Diego than it was to drive from San Francisco to San Jose during rush hour. Becasue he had skilled remote hands, he didn't have to fly down often.

    Some thoughts on networking for colo providers:

    • I used to think that I wanted to have servers closer to the exchange points. Exchange points are crowded places for traffic. If something happens elsewhere on a network, the traffic that is rerouted near your upstream routers might saturate your connectivity to anywhere. I'd now rather put my servers in a less-crowded area (St Louis? Denver? Dallas?) with good connectivity and let my bits get to most of the country/world without the congestion at the peering points.
    • A Tier 1 ISP may not be as good as a Tier 2 ISP with multiple transit paths into Tier 1 ISPs. I use SimpleNet in SanDiego as an example. They have six or more DS3/OC3 circuits out of their data center to different providers. Most other colo providers in our area have only one or two outbound paths. If connectivity out through one ISP starts to suck (fiber cut, BGP flapping), an ISP with multiple transit links can just shut off the bad ISP until it recovers.
    • If XX% of my traffic goes to a particular online service provider (AOL, CompuServe), a broadband network (@Home, RoadRunner), a WAP network (Sprint, others), through a portal (Yahoo, Excite), or to a dial-up audience (UUNET, AT&T), I'd want my servers to have really good conectivity to their network. I might choose a data center that has special peering or connectivity to my target network.
    • It really sucks when a colo provider oversells their capacity. Find out how much capacity a colo provider advertises. Then figure out how much 40% of that capacity is. Thell them you have BIG plans for a quick rollout. If they say they can handle it and don't suggest that they to order more capacity first, then they are likely to bump against their physical outbound limit which could mean packet loss for your site and every site withing the data center. You, already being a customer, would lose. If a colo puts you and everyone else behind bandwith limiters (like Xedia), that's good. They can rate-limit bandwidth hogs to keep everyone else running smoothly.

    Random musings:

    • One set of colo providers that I'm gaining respect for are the ones that try to do as little as possible. Imagine a climate-controlled U-Haul storage facility with padlocks in front of 10x10 rooms. Add conduits for cables that go to other cages that contain LEC fiber or ISPs' routers. Even a high school dropout can reliably run such a facility. I can take care of my own racks and networking. I can deal directly with my upstream ISP's without the colo middle man. Equinix follows this model.
    • Another type of colo is gaining my interest as well. Look at rackspace.com [rackspace.com]. Dell and Intel and perhaps others have the rent-a-server model. Sure, it's managed-server like others offer, but they keep themselves on the physical/parts side and can custom-build and install servers to my specs within a couple days of lead time. They don't try to manage the server at all - just rent me assembled server parts and bandwidth. I would no longer have to buy servers, just rent them. If there's a problem with a server, I can have them repair the server or build me a new one right away. I don't have to wait for my vendor to ship me a spare part to install - it's their server, and they damned well better have spare parts.
    • Distributing static content through a CDP like SandPiper or Akamai can significantly help the scalability of your web site. The only machines that you'd have to colocate would be site-critical back-end servers and masters for your site's content. A serious mistake that sites make is pumping too much bandwidth out of their colo provider. The CDPs have gigabits of good bandwidth all around the world to rent you. Try not to use more than, say, 20% of your colo provider's available bandwidth. As your site grows, your colo provider (with 4-6 months lead time on new fiber uplinks) might not be able to keep up with you if you depend too much on them for bandwidth.
    • If you have servers in a remote colo facility, like on another coast, make friends with a local geek or have a really smart consultant available near that facility to save yourself 5-hour plane trips. Frequent flyer miles don't make up for the time you waste on airplanes, and it can get old fast.
    • You can never have enough tools or spare parts in your colo facility.

    Just stuff to think about.
    --
    Eric Ziegast

    (PS: I used to work at ISPs and colo providers just like synx. I currently help run a very popular web site at several different colo facilities.)

  • I'd be a bit more cynical if snyx gave his company's name, web address, email address, etc.

    Which he didn't do.
    --

You can measure a programmer's perspective by noting his attitude on the continuing viability of FORTRAN. -- Alan Perlis

Working...