Slashdot videos: Now with more Slashdot!
From memory (I've been out of that business for 6 months) CPanel stores mail as maildirs. If you have gazillions of small files (that's a lot of email) then XFS handles it a lot better than ext3 - I've never benchmarked XFS against ext4. Back in the day, it also dealt with quotas more efficiently than ext2/3, but I really doubt that is a problem nowadays.
If you aren't handling gazillions of files, I'd be tempted to stick to ext3 or ext4 - just because it's more common and well known, not because it is necessarily the most efficient. When your server goes down, you'll quickly find advice on how to restore ext3 filesystems because gazillions of people have done it before. You will find less info about xfs (although it may be higher quality), just because it isn't as common.
XFS is probably better for large maildirs, but ext3 in recent kernels has much better performance on large directories starting in the late 2.6 kernels. It doesn't provide for infinite # of files per directory, but it doesn't take a huge hit listing e.g. 4k files in a directory anymore.
The comment would be purely theoretical problem.
Unless it can be seen in the binary, RTS will tell anyone involved: 'No, you cannot see our source. You've made a serious public accusation. Do you own your house? Any other assets? What was your net worth prior to today?'
If I understand your comment, you're stating that, if I may paraphrase, "of course they're not going to say they infringe. anyone would say that and since they only ship a binary they can keep the source to themselves while threatening legal action".
Again, if I understood you correctly, then you're also expressing that in your view they are guilty until proven innocent. I'm commenting since that's not the case. If I write some code and then release it under the GPL, then as Nick says in his comments, it is any other user who is responsible for abiding by the GPL. I, however, can chose to release my code in any form I chose under any other license I chose subsequently unless/until I assign the rights to another party (this is what the FSF asks for when you work on their projects - this prevents software that they advocate from being dual-licensed and allows them to unambiguously pursue infringement).
True, that. In '92 compuserve was established, but its greatest value for geeks was that they had a usenet feed and a mail gateway (which was probably a uucp connection to uunet/alternet, but mail flowed!), and so you could communicate with the rest of the world. It's still sad that they kept denying that this was their future until they couldn't stop hemorrhaging users.
Yes, the cell carriers will have a disruptive change hit them at some point, though. Their pricing is exorbitant and can't be sustained.
My point is that those "bunch of tiny servers" vastly outnumber the "real enterprise applications".
Shouldn't your hosting provider be doing this for you, or shouldn't you be doing this on install? This thread has been going on and on over a handful of config options. So you need configuration management. The apache config is flexible enough and unlike sendmail completely readable and comprehensible.
> I don't see that Oracle has anything to lose here by staying open with that component, filesystems benefit a lot from widespread use and lots of testing, but, well, it is Oracle.
I believe netapp still believes, somehow, that zfs is wafl, and that they should be paid damages for distribution of their IP.
I know that Daniel Philips has claimed at conferences way back when that he has seen prior art on WAFLs patents, but he still stopped working on Tux2 instead of fighting it. I don't know if Larry and the big O have a patent portfolio that can shut up netapp.
ZFS has been riddled with bugs in practice. Production crashes, re-silvering that fails constantly, magic voodoo incantations to get all pools up and running (I don't mean commands, I mean "well, sometimes the third time we reboot it works"). Uggh. Now that bugs aren't being fixed in opensolaris, I don't know how paying customers can convince solaris support to patch bugs. I used to point out that we wouldn't be paying for a patch that someone had integrated into opensolaris a year prior. I am so happy I don't touch this any more.
They're OK... until you try to manage different (commercial) applications on them. When app 1 requires a kernel patch, well there's no real virtualization there - the zones still run the same kernel, so when app 2 requires a different, incompatible patch, you get the throw up your hands and become the IT that says "no".
These are old issues, but trying to sell zones as the end-all be-all, or as even much more interesting than a BSD jail, is bogus.
Let's get to real issues that this doesn't change: patch management is a nightmare on solaris. 11 hasn't changed this. The OS is waaaay overpriced vs. the competition, and very unsophisticated processes monitoring via smf (I honestly think they should have cut their losses and just used runit - most of the benefits, none of the academically-inspired and simply stupid limitations in compiling the graph at boot time vs run time vs build time.... ugh).
Sorry, but no. I've been in the industry for over 10 years, and it's rare to experience these types of problems. Consumer-grade equipment is notorious for this type of thing, but it's much less common with a major vendor with a reputation to protect. Single vendor rarely means best of breed.
It may be rare, but when you're in a conf. call with Juniper and Cisco and F5 because you're finding that multicast is dropping packets, you can be pretty sure that the one that fixes it is the one who has a proposal to replace all of the others' equipment with their own.
A big part of Feynman's genius was his ability to explain. There are few people who could explain concepts like him, and so it's hard to get the education he gave in his books.
That is the hardest part to figure out. Current-generation LEDs have a tendency to flake out because of heat is my understanding.
This is a soft-sell way to get nosql databases into traditional IT situations, where familiarity with SQL will let current support and DBAs say "oh, it's like SQL, but it doesn't have joins. I can do that".
I always did like the sqlite docs, specifically the diagrams of the state machine for each statement.