Nothing says "You've been missed" when after 2 days off, you've 1500 emails, 3 'really critical things' and a 2 day backlog of work.
Including one absolutely fookin' critical restore, an urgent demand for additional disk space on a veritas cluster, and a pointy haired manager demanding I rename servers to 'conform'.
I'm considering my opinion on that latter. Ya see, the naming convention is basically just a bag of spooge. Basically, all our servers end up as S?RUGnnnn where the question mark is to signify the 'role' of the server. Unfortunately, the 'spec' was written by an arsehat who gave a load of letters, the only one of which is typically relevant is 'A' for 'Application'. So we have an awful lot of SARUGnnnnn servers, which is starting to get chronically meaningless. Increasingly of late, there's been queries of "can you check so-and-so on SARUGnnnn' which gets the response 'which one's that then?'.
(To which I typically interject 'I told you so').
Now I'm being told that "When we have the single console it will be important to be seen to conform.", which frankly is just bollocks. Yes, I fully agree that there are situations where you want similar names. When you've a cluster of 500 nearly identical servers for example. But we're rapidly reaching the point where we have 'generic server number 1, 2, 3, 4, 5,' and so we have to go look it up in our database in order to have any idea of what it is/does/who's it is etc.
When one is doing systems admin, getting your servers mixed up is really not a good thing.
Including one absolutely fookin' critical restore, an urgent demand for additional disk space on a veritas cluster, and a pointy haired manager demanding I rename servers to 'conform'.
I'm considering my opinion on that latter. Ya see, the naming convention is basically just a bag of spooge. Basically, all our servers end up as S?RUGnnnn where the question mark is to signify the 'role' of the server. Unfortunately, the 'spec' was written by an arsehat who gave a load of letters, the only one of which is typically relevant is 'A' for 'Application'. So we have an awful lot of SARUGnnnnn servers, which is starting to get chronically meaningless. Increasingly of late, there's been queries of "can you check so-and-so on SARUGnnnn' which gets the response 'which one's that then?'.
(To which I typically interject 'I told you so').
Now I'm being told that "When we have the single console it will be important to be seen to conform.", which frankly is just bollocks. Yes, I fully agree that there are situations where you want similar names. When you've a cluster of 500 nearly identical servers for example. But we're rapidly reaching the point where we have 'generic server number 1, 2, 3, 4, 5,' and so we have to go look it up in our database in order to have any idea of what it is/does/who's it is etc.
When one is doing systems admin, getting your servers mixed up is really not a good thing.
no subject
Date: 2004-11-09 07:29 am (UTC)I believe their standards come in four parts:
The type : Resource, Domain controller, etc.
A subtype? as above
A location name/company name
The local number (which could contain a subtype)
So you could have
Res/dcr/location/number
Sqm/rs/location/number
or
even
Res/rs/location/number
i.e.
APPRSLTSBDCnnnn
Private Banking still retains its model which is far easier to understand and usually describes the main function of the server.
location/function/number(if required)
i.e.
BHOffice - The office server for Birmingham
BHDev - Development Server
Simple vs 'This page left intentionally blank'. I know which one I would choose.
no subject
Date: 2004-11-09 09:08 am (UTC)ARRRGH!!! The Nightmares, The Nightmares!
Yeah, in my old job we had to deal wtih that naming convention a lot.
I just came to the conclusion that IBM UK has about 3 half compitent people in it. And they're either leaving or getting fired for not brown nosing enough...
no subject
Date: 2004-11-09 10:35 am (UTC)no subject
Date: 2004-11-09 12:26 pm (UTC)no subject
Date: 2004-11-10 08:54 am (UTC)no subject
Date: 2004-11-10 08:57 am (UTC)SARUG10332
SARUG10323
SARUG10233