which is probably why i like it :freak3:

7/10 - Must try harder :)

"Planned" outage ;)


At a much broader level this would be encompassed by business continuity management. Refer to BS25999 :)

Mmmm Im a solaris contractor, and normally a situation like this would be called a sudden system outage / restart, as its not just a simple case of rebooting servers, especially if the san / fabric has gone down. You need to implement a structured restart.

Mmm all sounds good in the real world, but with my current employer, this situation is referred to as "A good old fashioned witchhunt". Dont worry about fixing the problem or rebooting, no, you need to follow the below steps.

1) First things first, you need to sharpen the ends on your pitch fork, and hunt that man down! Look around your office, the person last in, or who hasnt turned up is always a good start. On a weekend or late night, the person who had his phone off, avoiding the whole thing should always be blamed ! as why should he enjoy his life when your stuck in a cold server room, on a Saturday night !!! Or failing that, last one logged on must of done it.

2) Once this 8astard is found, you need to keep telling him where he went wrong, keep going on about why it shouldnt of happened and express how much better you are than he is. Rolling of eyes, abusive muttering and mentioned of P45 should be implemented. If a contractor, repeatively mention who contractors are just money grabbing c*nts !

3) Then, move on to a conference call. Try to have as many non technically people on the call, make sure they all love the sound of their own voices, and limited the number of technical people, who are actually fixing the problem, to just one. More points are rewarded to indian people with heavy accents, people on mobiles, people on mobiles on the train. Dont talk in an ordered manner, all talk at once, contradicting each other, slagging off each others opinions and pretend to know what your listening to with the odd "Mmmm yes I agree".

4) Next, from the call, implement the most stupid, the most unrealistic, the most pointless action possible, suggested during the call. Never listen to the technical guy, his years of experience are worthless in a situation like this. No, clearly the person with the loudest voice, or most agressive nature should lead the call.

5) Once the action has been implemented, and becomes obivous that it hasnt worked, got back to step 1.


So, perhaps instead of `DR`, this situation should be know as "Working in a Gov. institution".

SBK

Updated term is Continuity Management HTH ;)

BTDT - 3.30am, on a mobile to Bangalore talking to a guy named "Love" about a broken MySQL database that had reduced all of a certain "e-booking" site to a standstill who, on top of a thick Indian accent, also had a speech impediment.

Sorry, it isn't.

Continuity Management is business w*nkspeak for any kind of outage over any area of business.

In IT terms, this is called DR.

I hope that helps.

I might add, the act of bringing service levels back up, as in the original post, is still "DR". Any plans put in place, before the business starts trading ideally, is part of CM, which will include IT's DR strategy.

I had a conf call on Friday with a spanish guy who must of learnt his English at some dock yard, a indian guy who's accent was so thick I couldnt understand a single word, and an opinionate fat little woman who hates men, plus 25 others. Pure hell.

SBK

Contractor definition for `DR ` = A situation where one can charge endlessly, morning, noon or night, the contractor till keeps ringing :-), otherwise known as "Milking the financial cow".

SBK

No you are very much mistaken there. In that situation the contractor has to come out of hiding and is required to do work and become noticed. Not good.

Cotton and MattW, I reckon you two need to sort this out via conference call.

SBK

Gaggle pin reset.

FUBAR

I've just spent 4 hours on a conf call because one of our "server/network things" screwed up due to a power failure. With 1 sites set of servers + routers in various state of dead/config loss coma.

The built-in fault tolerance allowed the second site to seamlessly take the load. We certainly didnt invoke "Disaster Recovery" because as far we could tell this was a simple power issue and not actually a Jumbo jet hitting the site. Full DR as I know it would have involved new servers, a new site and some serious management panic.

Techinally "it was fubar" and "we fixed it". Fortunately in this instance 3 techies managed to get it done before the headless management instigated sbk1972's very accurate description of a conf call.

I was on a call once when a high profile national government system was down. We had the required 10:1 management:techy ratio. They delayed service restoration by over an hour as they preffered to discuss the impact of not fixing it, rather than get on and fix it

Lol very accurate. We have so many support teams at work now to tell you that the fault your working on has high visibility. How about instead of wasting my time with escalations you do one and let me fix it? Better still sack you and use you head count as a tech rather than pencil pusher :)

Simon