[AC-Admins] Trash talking servers / admins

[Pippin Bear]

On Sun, Feb 03, 2008 at 02:33:19AM -0500, Scott 'Simba' Garron wrote:
> 06:04PM <Inaki> lion and snowleopard are garbage

Not really true.  lion used to have problems, as you mentioned, Simba.
I do still notice links to lion dropping occasionally and then coming
back, and it's probably my memory of you having occasional complete
(though perhaps short) network outages (problems with your external lines
as far as I remember from the time) which may have been a few years ago
now that's making me think it's the same thing happening now.

Snowleopard's been pretty solid up until the last week or three.
I suppose really it ought to have been taken out of DNS rotation from
before the move until it had been bedded in for a couple of weeks,
then added back in later.  Then the kinks could have been ironed out
without the whole network seeing regular netsplits.  We do (did, when
we had new servers) that with new servers, so we should probably do the
same with moved servers.

> 06:07PM <Inaki> Lion's been okay recently, but ya.. it had all sorts of 
> problems
> 06:08PM <Pippin> lion is on a network whose uplinks tend to have outages 
> sometimes.  lion itself is stable as far as I know, but with a crummy 
> network connection and services hosted with it, it tends to disrupt the 
> whole irc network if it hiccups.

Sorry, I shouldn't have said "crummy".  That was likely the combination of
(a) memory of some network outages around lion, and (b) lion being the
network hub, and thus any outages having greater impact on the network.

> 06:09PM <Inaki> also trolls have root on that box >.>

This I have no idea about.  I almost asked at the time, but thought I'd
keep my yap shut.

I'd agree it might be well worth while starting up those staff meetings
again.  If we have a time we know we have to talk about the network
for a bit, maybe we'll mention these things that have undoubtedly been
swilling around in our heads with no way out.  :p

> (those minutes are very few and far between.  Kudos to Pippin for having 
> stellar uptime with that).

If you mean nagios, that's hosted on my rented box in germany and has
actually had more downtime than I'd like.  They took nearly a day to
replace a brokened PSU one time, and their network is a bit shoddy,
which is a definite source of false positives on nagios.  I hope to
move nagios to a box in my own rack sometime - I want to get rid of that
rental machine and keep the £40/month I'm currently paying for it!

>       Now...   With those facts, what's really on your minds?  If 

Dunno. ;)

Pippin