Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
For the out-of-band Slashdot experience (mostly headlines), follow us on Twitter, or Facebook. ×

Comment: Re: Outage.. (Score 1) 334 334

"They are trying to find someone with the exact skill sets and pay them as much as the guy who left for a better job. So they let a good resource leave, and they haven't learned from their mistakes and either realize that they will need to lower the requirements, or raise the salary and benefits."

Yes.

Going back to the first post on this thread, all this means that, in a company, "the service" is much deeper and wider than thought at first glance and, say, a server breaking can have its root causes very far away from the server room.

It pays to have an holistic view about the business, but very few companies pay attention to that or are even organized to facilitate such a way.

Comment: Re:That's still exactly what it was (Score 1) 201 201

"Arguing that Hitler wasn't compassionate about his precious Arian race. Not on good terms with reality eh?"

He was so compassionate that he didn't doubt at putting as many arians in front of their enemy's gun machines to be killed as he could, including women and children and stating that if "his" people couldn't win the war, then they weren't worth to exist (i.e. not surrending after carpet bombing their big cities, using children for the last stand of Berlin...).

No, there's no need to be compassionate to be a great leader and, sadly, in so many cases, it is the sociopath the best one, since he knows what to say to whom in order to make them move, disregarding the consequencies.

Comment: Re: Outage.. (Score 1) 334 334

"I just read you as, "not my problem""

Yes, that's the case... from a certain point of view.

I usually respect enough others' work as to give them their due credit. In this case, it means I credit the system architect as being able to design the system properly. No high availability means it's not a critical server, so I adapt my procedures accordingly.

"figuring out what went wrong is precisely your job"

No, it isn't. My job is to produce the most value for the company within my assigned competencies. Sometimes it means scratch my head for hours to solve a problem. Some others it means reboot/destroy a server wihtout a second look then go to the next item on my to-do list. You know, servers are not pets but cattle.

"You sound like a dick answering a different question than asked"

In fact, I didn't. I was asked to solve the problem, not to diagnose the problem and solve it without rebooting the server, and I honestly gave the answer I considered to be the most effective. As it resulted, it was not the answer my interviewer expected nor wanted but I'm fine with that: in a hiring process the prospective employee is interviewing the employer just as much as the other way around.

Comment: Re:Outage.. (Score 3, Interesting) 334 334

"As with most mistakes, it is part of a system that is faulty and awaiting one simple mistake to escalate."

Can't agree any more.

"Chances are there was a culture of trying to save money"

Sometimes the "cargo cult" is so ingrained that even the techs are unable to see it.

Anecdote:

Was in a hiring process, not remember if it was Google or Amazon. One of the questions (from a hands-on tech team lead) was about a single server that went crazy and couldn't spawn any more processes, so it was almost impossible to do nothing with the computer. It still was offering whatever services it hosted just OK.

It went more or less like this:
Me: Has this happened before?
Recruiter: Nope.
Me: So... Can I try this, or that, or this other one?
R: No, because you can't run any new process.
M: Ok, reboot it (I of course know saying somehting like that is taboo for a unix/linux sysadmin). Let's look at the booting messages to see if we get some clue and let's monitor it afterwards to see if this happens again. If that's the case, we will be in better position to diagnose, if not, we will put it on the "computer gnomes" account.
R: Won't try to diagnose anymore before rebooting?
M: Nope. My time is valuable and there will surely be more productive things on my to-do list.
R: But the computer host a service that if turned off will cost the company a bazillion!
M: Nope. If that were the case, the powers-that-be would have engineered the service with high avaliability in mind -which in turn means we could reboot the server without further hesitation. Since that's not the case, the implicit is that business already considered it not a critical service so point above about me costing money still applies.
R: But, but, but...
[...]

Of course, I knew from the very begining the answer he wanted was to find a way to list the process list without spawning a new process so after a while I went throw that route -I vaguely remember there was some Bash built-in that would allow me to do it, but not exactly which one, but back in that time I wanted to see the culture of that place.

There's no need to say I wasn't hired. But I didn't wanted to be hired either. Not within that team at least.

Comment: Re:Profit over safety (Score 1) 127 127

"no one was under time pressure to drill what wasn't even a production well."

Yeah... only the date to finish the works on it was March 8th and any delay past this date was incurring overruning and opportunity costs since the Deepwater Horizon was already assigned to a new project.

Accident was on April 20th. These 43 days costed around 21M US$ to BP. The very same day of the accident a recommended 9-12 hours and 128,000US$ test of cement status was cancelled.

"Now please educate yourself beyond the daily telegraph."

It is not the messenger the one that makes the message true or false but, anyway, do you find the U.S. Government Publishing Office to be more reliable? Ok, try here for the same cite: http://www.gpo.gov/fdsys/pkg/C...

So the US Goverment publicly states that: "Whether purposeful or not, many of the risk-enhancing decisions that BP, Halliburton, and Transocean made saved those companies significant time (and money)."

Maybe its some thegarbzs overthere the ones that should better educate themselves before asking others for the same.

Comment: Re:Terrible. (Score 4, Insightful) 197 197

"When an act of vandalism affects many thousands of people, it's pretty fucking serious."

It depends on how those many thousands are impacted. When biggest impact is not being able to download kitty videos at full speed, no, it is not "terrible". "Annoying" comes to mind, but not "terrible".

"Many people and businesses have their telephone service over the fibre"

Even if we accord at calling this "severe", severily impacting business is, well, "severe" at most, still not "terrible".

Adjective inflation lets you without terms for really big things. If "people may notice slower email or videos not playing, but may not have service completely disrupted" is "a terrible crime" what does this leave to, say, 09/11attacks? "a terrible crime indeed, I really mean it"?

Comment: Re:Profit over safety (Score 1) 127 127

"You mean an oil spill which was the result of engineering, judgement and training errors of the operators and had absolutely nothing to do with time pressures"

Uhhh... nope.

I'm talking about oil spills that even the government comission in charge said that were due to cost-cutting malpractices, i.e. "Whether purposeful or not, many of the decisions that BP, Halliburton, and Transocean made that increased the risk of the Macondo blowout clearly saved those companies significant time (and money)": http://www.telegraph.co.uk/fin.... While the report itself can't be so clear-cut on that, it's very easy to read between lines that conflicting interests (time and money versus security) just as stated. http://ccrm.berkeley.edu/pdfs_...

"Or the nuclear power plant which survived the tsunami just fine [...] and only went under due to the engineering error of putting emergency power in the basement?"

Uhhh... nope again.

I'm talking about nuclear power plants that were positively known not to be properly maintained and where whistleblowers were systematically shut down because hearing them would cost money, i.e. "It is important to remember that in February 2011, shortly before the meltdowns, NISA extended the operating license of Fukushima Daiichi despite expressing reservations about a dubious maintenance record and eerily prescient concerns about stress cracks in the back-up diesel generators that left them vulnerable to inundation." or "Telltale warnings began accumulating over the decade prior to 3/11[...] In 2009 NISA and TEPCO discussed the possibility of a 9.2 meter tsunami based on new simulations and archaeological evidence, but NISA did not press TEPCO to take countermeasures. Clearly, there is no basis to TEPCO's claim that the scale of the 3/11 tsunami was inconceivable [...] In terms of tsunami-related risk management, it turns out that TEPCO and two other utilities actually lobbied the government's Earthquake Research Committee on March 3, 2011 to water down wording in a report warning that a massive tsunami could hit the Tohoku coast". http://www.japanfocus.org/-Jef...

"Please if you're going cite major disasters to support your case for management attempting to maximise profit at the expense of safety then at least cite some correct ones."

I did.

Comment: Re:Wrong about automation and profit (Score 1) 127 127

"The article carries echoes of the "profit is evil and government is good" mantra so popular lately."

Is it that, or might it be "there are business where it's probably better not to center the whole focus on profits, which is what for-profit companies are forced to do"?

Of course, your mantra is shorter and more black&white-ish, so it must be true.

Comment: Re:Profit over safety (Score 1) 127 127

"These kinds of things are (fortunately for us) regulated."

That's why, say, an oil spill tied to the need to open a platform ASAP will never happen. Or... a nuclear facility will never suffer a melt down because management prefers to turn a deaf ear to the researchers that say more expense against tsunamis is needed.

Oh, wait!

Comment: Re:Just private contractors? (Score 2) 127 127

"He thinks it's just private contractors that cut corners to save money? "

While government *may* cut corners to save money, since their percived profit not necessarily is tied to pure monetary profit, private contractors, *must* cut corners to save money, since their profit is pure monetary profit.

Comment: Re:Comment (Score 4, Insightful) 80 80

"But what I'm getting at is when the government pays employees way above the per capita income of taxpayers."

I see your point, but I think is moot both in practical and ethical grounds because tax-to-wages is not a one-to-one map.

In practical, it's just a market, offer and demand: if you want the best (not that you *must* want them, maybe your needs are not up to the best and brigthest but *if* you need them), you need to have three things:

1) Have a selection process to filter out everybody but the best.
2) Have a process in place to detect the mistakes on point 1 above and/or those that, still being the best when hired, may be not the best now.
3) Offer the highest pay (not necessarily just money, other perks included) so the best apply to your selection process to start with.

And with regards to ethics, I don't see how someone can be comfortable paying, say, a bartender the same than to a, say, brilliant doctorate in something you really feel useful. It is not as if that any single doctor needs to be payed in full by any single bartender's taxes. It is still reasonable for an hypothetical town of "humble farmers" to pay their, say, doctor, or judge, or sheriff, above their own average income if they feel their value to the community is also above their own average.

Comment: Re:Comment (Score 1) 80 80

"If everyone wants to get paid above average, half of the available positions will remain unfilled."

Only if there's something like 'basic rent' in place.

In the meantime, in the real world, it's not so much about what you, as an employee, want to be paid, but about what you, as an employer, want from your employees. And then, if you pay peanuts, you get monkeys.

Every program is a part of some other program, and rarely fits.

Working...