« first day (856 days earlier)      last day (4411 days later) » 

19:00
Such anger.... You people suffer from a serious rum deficiency.
@ScottPack When I'm work, I do.
@HopelessN00b I hope the wiki tag work cheers you up. You're quite the busy worker!
@ScottPack But you'd be angry too, if you go the tickets I get.
@HopelessN00b Probably. The majority of my tickets are for approvals. Yes...Yes...Yes...install updates first...yes...more information...MOAR INFORMATION...
@pauska Honestly... it's something I can without a change management meeting, a flurry of ignorant, dumbassed objections and 3 layers of red tape. Most infuriating thing about this job. All the damned paperwork, no productivity.
19:03
@HopelessN00b sounds like you need a new job.. don't work in management if all you really want to do is tech stuff
@HopelessN00b Hey, we have a couple of openings.
@HopelessN00b Is it $$$ or just $$
@JoelESalas nope @HopelessN00b it really, really sucks
Hey ya'll... quick poll on KVM switches and preferences (what do you prefer?). We have a belkin OnmiView Pro3 that has caused nothing but problems. Trying to find something more reliable.
The guy who ruined it for us has his own office now, so he can listen to his music and now we're stuck without it
Fucker
19:09
@Cole What is the motive behind not letting people listen to music
@pauska Yeah, I'm not in management. I'm allegedly an SA... and I guess I am, but we just move soooo slow.
@JoelESalas Just $. Average wages, average benefits. Counting the days until this infra project is done, lemme tell you.
@JoelESalas this guy had his headphones in and the Director called him 6 times and he wasn't paying attention to his phone. Director came over and stood in his cube door, and the guy turned around all "what?"
Director flipped out
That's why we can't wear headphones.
Sounds like somebody has a bad case of the sandy wide set vaginas.
@ScottPack Lol'd
@ScottPack Yeah, but I doubt you can call him on it. One thing I've noticed is that the amount of explosive rage someone exhibits over being told to knock the sand out his vagina is exponentially proportional to the amount of sand in there.
Like calling someone an asshat. The more asshatted they are, the worse they react to being informed of what everyone else knows.
19:14
@Cole well, that is kind of understandable.. I'd probably get mega pissed aswell
we're allowed to use headphones here, but not so loud that you can't hear your phone
@pauska agreed - but now, that guy, has his own office
Never use e-mail when it's time sensitive. Never use a phone when you can just walk over and talk to the person. Don't use technology as a crutch, or a way to avoid people/human contact.
so he can listen to music and what not, and now we're stuck without being able to wear headphones
@ChrisS i occasionally gchat people 20 feet away from me
@ChrisS there's a dilbert for that
19:16
@pauska there's an app for that
@pauska lol.
@ChrisS I'm confused. If we're not supposed to use technology to avoid contact with the users, what the hell good is technology?
@HopelessN00b It's supposed to make you more efficient.
Here's a question...
19:20
@ChrisS So is avoiding direct contact with the lusers. :p
Client wants disaffinity rules for their ESXi cluster.
@ChrisS I used to think that until I started annoying the shit out of my coworkers by walking up to their desk.
e.g. app01, app02, app03...app06 on different nodes
however, they only have four ESXi hosts.
how would you arrange the app servers?
Put them all on one ESXi host and when they complain about performance sell them more hardware. =]
use vsphere 5, and use groups with "should run on host xyz"
19:23
@pauska Stuck on version 4.1 here.
I have DRS...
can't remember if 4.1 has that option
but when I have 6 app servers and 4 hosts...
3 app servers on each host
err 4 hosts.. um
@pauska Yeah, that could work. Arrange it like a RAID5 array. Just, instead of disks that will ultimately fail if you ever need the redundancy, you'll have VMs that will ultimately not failover if they ever need to.
I usually do the rules in pairs... to make sure pr0n-app1 and pr0n-app2 are on different hosts.
19:25
well, figure out which app server talks the most to which app server
well, these are pubic-facing systems...
are the six different app servers all in one tier?
@ewwhite Is this really Brazzers again? You gotta do something to convince them to embrace redundancy. I don't know what I'd do if my pr0n went down.
yes
@HopelessN00b ;)
@HopelessN00b no, the drs rule "should run on host" fixes that.. if HA/maintenance kicks in they get placed wherever theres room
@ewwhite but.. what does it matter if they're on the same host? Isn't it enough to guarantee that not all of them are on one host?
19:27
checking to see what this client actually does...
like 2+2+1+1?
ugh headache out of NO WHERE
@pauska Yeah, like how RAID5's rebuild kicks in on a hot spare and saves your array auto-magically if a disk fails. Works better in theory than practice. (Though I will admit that VMWare's DRS works a lot more often than a RAID5 rebuild.)
if they're on the same tier and load-balanced properly it really does not matter
"I’d like to put some protection policies in place in reagards to our esx farm. Sunday we lost an esx node that had too many vm’s on it and had pairs of vm’s that shouldn’t have been on the same esx node."
In my opinion, we shouldn't have lost an ESXi node...
:-X
19:29
@HopelessN00b I'm sorry man, but you're way off right now.. You're talking about fault tolerance. I don't think ewwhite's porn site can afford that backend :)
ahh
actually, this is not a porn one!
well just ask him which vm's that shouldnt be on the same host
it's just one rule and a big group in DRS..
we do that with lots of stuff..
@ewwhite I think the first step is finding out why these two guests shouldn't be on the same host. It might be much ado about nothing.
err no sorry, one rule per pair
@HopelessN00b probably because they are sitting behind a load balancer.
and losing the host made the entire app go bye-bye
19:30
@HopelessN00b it's not that uncommon.. two exchange servers (when all you have is two), your sql mirror pair and so on..
@ewwhite Well, that's a pretty good reason, then.
(granted, the host should have had redundant power supplies...)
@ewwhite agreed, but still no excuse for not having proper DRS rules :-)
Well, anyway, I'm off to a remote site. To tell them why they can't just move a rack of gear to the other side of the building. <sigh> Later, all.
we're going to need our own meme-wiki

big endian
Pauska's Sock
Pauskasock
ewwhite's porn server
dancing b*witched bandmembers
Tomtom's phone servers...

oh my!
19:32
hahaha big endian
forgot about that one
wasn't it Bart? he got so angry at us
@pauska it seems easier to do this in pairs... eg. app1 and app2 can't be together. But what is the group is 6 servers?
on 4 hours.
@ewwhite a 6 server group does not make any sense
What this Big Endian stuff??
I just connected tomtom with my old boss in Germany
@ewwhite you must hate your old boss
@ChrisS I can't remember the details right now.. it was someone here talking to one of the girl blue's about something
19:35
@pauska They have the same personality
then suddenly he asked her, "do you have a big endian" or something in that order
quite easily misread, and so it began..
Just like TomTom, my old boss can be very nice and charming outside of a SF or working context.
@ewwhite anyways.. 6 servers with total resillience does not make any sense, unless they have unique functions/datasets..
ask them how many app servers they can lose..
@pauska I think it was @WesleyDavid and rebecca chernoff
And yeah he was a little embarassed
19:42
@RobM I wasn't embarrassed in the least.
of course not. Kittehs have two giant brass ones
She was doing her backwards smilies and I asked her if she was big endian or little endian. Everyone else took it and ran.
ah yes, that's it.
I'm still waiting for her answer.
maybe you should remind her about the question.
It wouldn't be serverfault chat if we didn't take at least one thing a day out of context
19:44
Where is The Chernoff anyway? Haven't seen here in a long time. =(
So I'm going to set up a LinkedIn profile to bait recruiters. I'll say I've got experience with SAS, DAS, NAS, MAAS, SAAS, GrASS and BASS.
@tombull89 DAACS
Devops as a Cloud Service
4
@WesleyDavid 3rd from last on the picture wall
And when they ask what GrASS is (because I've made it up) I'll answer with "Grass. You know, the green stuff in a field? Jesus, you should get out more".
I have DaaS, but only for the ladies
19:45
devops as a cloud service. Now I feel ill.
ahhh it was in the election room
in Server Fault Election: Town Hall Chat, Jan 16 '12 at 21:52, by WesleyDavid
@RebeccaChernoff Are you little endian or big endian? That would explain the smileys.
that's the one
@pauska asking now
And don't y'all want to know which client this is?
@HopelessN00b Here, go to the kitty cam right now. You'll feel better.
@RobM Here's an in the cloud web browser. How does that feel? commonit.com/en/products/airship
Noooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
19:56
@ScottPack Yay! It's feeding time!
@jscott inorite
I also want to fall asleep while eating. Looks extremely comfy.
Wow, not the response I expected. I went to my coworker, he's our "VMware guy" to ask where I could create a Syslog server.
@pauska No doubt.
Or, well, fall asleep while drinking would probably be a more accurate description, but we've all been there..
19:59
@Cole "What is a syslog?"
His response was so excited "THAT WOULD BE GREAT!" I responded with "wait was that sarcastic?"
He's like "No! I'm really excited!"
haha
@cole I can see it from his POV aswell.. getting a fancy syslog server up with web interface and what not is something all sysadmins would love to have, but never have the time to set up
or just a unix one that you can grep
vmware's internal syslog server is HORRIBLE
Yeah well I'm doing Unix based
It's not that they dont have time
they just dont know how to here.
windows shop?
@Cole Well...yeah. syslogd on Windows would be like Apache on Windows.
20:02
@ScottPack logstash
@ScottPack that's not so bad actually..
Syslog is actually pretty simple
It's just a directory on a unix server you can look through that shows you all the syslog events from the devices configured to come through it
0
Q: monitor/stream bash_history

Nikki Wilsonis there a way to stream/monitor/log bash history to prevent the ability of someone hiding stuff? would like to monitor and know every little thing that happens so that the list of commands and edited files can be checked and watched. Would rsync be an ideal way or is there an actual program th...

His answer...
@mossy haha.. what the hell
flagged tho, thanks
Lol
@Cole thanks for bringing this up btw - just found out that I've forgotten to install syslog for our vsphere cluster :|
PCI DSS is such a pain
@pauska haha
@RyanRies Yes, yes it is. Let's get drunk and be bitter together.
@WesleyDavid Aye, except for instead of a worm at the bottom of the tequila bottle, it'll be an RSA token
20:10
@RyanRies You think that's bad? Try out HIPAA/HI-TECH
@ScottPack I've gotta deal with it all man. We have both medical services customers and federal customers... FISMA/FedRAMP, you name it
@RyanRies I can relate.
Alright that's two drinking buddies then :)
my coworker is kicking the back of her desk
going to lose my shit
MY RETURN IS NEARING.
WHO IS EXCITED?
20:16
@RyanRies You're too far away.
@ScottPack I mean like, in spirit, man...
my office smells like B.O.
@ScottPack What do you know about HIPAA?
Starting to learn
@ewwhite When do you discuss pricing when you're working on a project with a customer?
Is it right after the specs are ironed out?
20:36
@JoelESalas Once you're sure they're not going to fire you unless you're grossly overpriced
@MarkHenderson This is a former boss, we have a contract in place. I'm just trying to estimate the amount of time, don't want to cause sticker shock
@JoelESalas I'm in a unique situation... people don't typically shop me.
Because my HW prices are pretty much the lowest.
I ALWAYS have this problem with CentOS
Whenever I configure eth0, I can ping the gateway - but I can't route out. I've disable iptables and I've put the gateway info in as well. arg
@Cole Disable the NetworkManager service.
@ewwhite is it under something other than "NetworkManager" ?
20:48
hi all
Because I do not have that on my box
service NetworkManager stop
Yeah NetworkManager: No such file or directory
hmm, okay
@ewwhite weird, right?
I can ping the gateway
20:50
a little
so whats a good backup?
@NickM. for...?
Need moar info
Having tape backup or having a another drive backup
But what are you backing up
@NickM. Depends on what your needs are
20:51
by another drive.. i dont mean raid.. i mean a separate unit just acting as a backup storage array
@Cole what does route show?
There's a bazillion backup products. Something that is a "good backup product" for me, may not work for you at all.
Are you looking for long-term archival? Then you want tapes. Are you after something quick to access so you can restore the secretaries FY13 budget? Then you're probably better of with disk
arg its in a VM so I cant even copy and paste @JoelESalas
There's no one size fits all and it depends on your needs
There are people whos entire job is just advising companies on backup solutions, it can be that much of a minefield
20:52
@Cole Do you have a default route?
10.2.0.0 * 255.255.0.0 U 0 0 0 eth0
@MarkHenderson 14 day retention on all data.. not talking about a specific product mostly asking if having a separate storage is sufficient enough or do people still do daily tape backups?
Did you configure that by hand
link-local * 255.255.0.0 U 1002 0 0 eth0
@JoelESalas yessir
@Cole this is what mine looks like
20:53
@NickM. Like I said, it depends on what you want and what you're backing up from.
[root@simulantproductions ~]# route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
default         gw-li165.linode 0.0.0.0         UG    0      0        0 eth0
link-local      *               255.255.0.0     U     1002   0        0 eth0
173.230.155.0   *               255.255.255.0   U     0      0        0 eth0
And what you're using the backups ofr
for
default connwebprod1.co 0.0.0.0 UG 0 0 0 eth0
Are you using them for disaster recovery? Accidental recovery? Auditing? Theres a million questions to ask
@JoelESalas mine looks right
For my config
network is a /16
20:54
Whats your budget? Are you under any legal obligations about where you store the data? Are you storing confidential data?
@ewwhite It sucks to comply with.
What MTR are you aiming for?
@MarkHenderson Just daily backups of all users home directories, databases and monthly full system backups
What kind volume are you backing up and what kind of bandwidth do you have to back it up with
@ewwhite This is just for labor, they already bought the gear (bummer)
20:55
@NickM. Then you need to think about this more, because how are you going to have 14 day retention on a monthly backup?
Jesus Corp AMEX card already arrived
Wife just got approval for new computer. She has a 6 year-old 20" iMac. She sees the 27" iMac online, but it's not available anywhere. There's a 21.5" iMac out there that has limited availability. Would it just make sense to get a Mac Mini and a 27" Thunderbolt display?
@JoelESalas I tell people what it will cost. I don't alter my rate.
@ewwhite Yeah they're hard to get apparently.
/flips desk fucking CentOS
@ewwhite why not just a MBP? + monitor
@MarkHenderson sorry im being too vague. What i meant is that user files and server database will be backed up daily with a 14 day retention. Also on this same storage unit, it will hold monthly images of the server
20:56
I guess that would be way too costly though.
@Cole For the record I've never had that problem (given a real working gateway)
@Cole she had a personal MBP... this is her work machine
@NickM. Are these Linux or Windows systems?
Windows
@JoelESalas what the crappola
20:57
And are you trying to recover from mistakes, or disasters?
and she's one of those people who uses an iPad for personal computing
(e.g. deleted files, or the building blows up?)
@ewwhite ah, well the mini's aren't bad. I was looking at getting one to replace my desktop at home.
@ewwhite I will heed this advice
@MarkHenderson essentially both
20:58
@NickM. Do you want to spend five figures on this solution
@Cole Let's see your ifcfg-eth0 file
@MarkHenderson I understand tapes provide that mobility however I'm really trying to stay away from tapes
@JoelESalas Clients love "not-to-exceed" quotes.
@JoelESalas alright give me a sec since I can't copy and paste -_-
@JoelESalas under 5 figures
@NickM. Because if you're looking to protect from a building explosion, then you want a remote, off-site backup. Whether that means you use a "cloud" backup, or ship tapes off-site, or even ship a 2TB external drive offsite, it makes your Mean Time to Recovery very large
20:59
Even if it's higher... a flat-rate all-inclusive number helps people sell your to the decision-makers.
But if you're looking to protect from dumb mistakes, then an on-site, fast, disk-based backup solution is what you probably wanty
Why the aversion to tapes?
@NickM. You might have some trouble acquiring enough storage. This is assuming you use FOSS to do the actual backing up
They're very expensive, but they have huge capacities and are very flexible with a ATL.
DEVICE="eth0"
HWADDR="blahblahblah"
ONBOOT="yes"
BOOTPROTO="static"
TYPE="Ethernet"
UUID="blahblahblah"
IPADDR=10.2.243.10
PREFIX=16
GATEWAY=10.2.243.1
@ewwhite I'll do that. Best I can tell it'll be 10 hours of work at most. Should I quote X, then bill X-0.1x ?
21:01
Oh you know
fuck me
@JoelESalas Just give them a number
@NickM. you probably want a combination of both. Some fast, disk-based backups locally to keep your local backups, and then once a month ship some full backups off-site. You can either ship them using cloud storage, an external hard drive, whatever
(you won't get tapes for < 10k)
goddammit! even when I outsource backups, they're still a pain in my ass
give a time estimate, too... but the not-to-exceed helps people plan.
@Cole That's not a valid HWADDR
21:01
@Cole Use NETMASK 255.255.0.0
@MarkHenderson The problem with tapes is that it requires having someone on-site. We do not have a tech on-site at that location to collect the daily tape backups and the users are as computer literate as a monkey
@cole use system-config-network
@NickM. Lol'd at the monkey part
@NickM. Well that's not just a reason to ignore tapes. thats also a reason to ignore all removable storage, and thus your only options for disater recovery are internet-based backups
Whether you have a VPN to another office, or a service like backblaze, you're limited for choice
@JoelESalas still didn't work @ewwhite trying that
21:03
@Cole Sorry, to clarify: Get rid of PREFIX statement entirely, use NETMASK instead
@JoelESalas Oh I did
@Cole Just making sure
This is quite a weird little problem you've got here
@JoelESalas what, do you think I'm a luser?
/flips desk
I ran into this problem with CentOS 6 before. Can't remember what the fuckity fuck it was.
@ewwhite Got it, will do. Thank you again for the advice
so iMac versus Mac Mini?
21:09
iMac
IPS displays are so nice
@MDMarra 4 week wait
Also, I want to play with the new Fusion drives
It would be the 27" LED Cinema display
You can't put them in the mini, right?
this is her work computer... yes, you can get SSD, 1TB SATA or 1TB Fusion on the Mini
21:10
@MarkHenderson thanks
@ewwhite I didn't know they offered it in the mini
is backup exec still the cool kid on the block?
I mean if you're splurging on a 27" Cinema anyway, might as well get the mini
cheaper to upgrade down the line and that screen will be quality for years
@NickM. netbackup
backupexec is it's red headed step child
Commvault and FalconStor both looked cool when I demo'd them too
But how excited can you really get about backups
@MDMarra @ewwhite even if she wanted a 27" iMac down the road - DUAL DISPLAYS!
21:16
@MDMarra I have one server to backup. Would netbackup be overkill?
@NickM. This is Windows?
server 2003
@NickM. Acronis is nice and cheap
will do mssql database as well?
@NickM. It's a separate add-on, but you're still talking less than a grand for a complete backup package, which is absurdly cheap
21:19
I'm with @JoelESalas
@JoelESalas Last time I heard talk like that, I ended up administering an HP DataProtector install.
I'm still traumatized from that.
Oh man I'm tired
What's the industry name for software that allows secure document sharing with outside partners?
There's a niche industry / software term for it and I forget. >_<
21:36
@WesleyDavid Fax?
@WesleyDavid DMS? Document Management System?
Found it. AS2 systems.
@WesleyDavid EDI?
Ok I'm going home, bbiab
@HopelessN00b AS2
Also, Managed File Transfer software (MFT)
21:47
Speaking of backups, does anyone have any idea how much Bacula Enterprise costs?
We're re-evaluating our VMWare backups, cos I'm getting the shits with Backup Exec
@MichaelHampton Bad Behavior is breaking my balls with oauth :<
Who was the person that said raid 6 was a bad idea?
22:04
Is it a bad idea?
@It's a bad idea if you have 4 disks, it's a good idea if you have more
use raid 2
@MattBear RAID 2: This time it's personal!
2
om nom nom Is there anything that Tabasco sauce can't make palatable?
<-- loathes double negatives
@WesleyDavid Diarrhea
@MarkHenderson What if it's Tabasco induced diarrhea?
22:14
SX GBIC versus LX GBIC? who can explain?
@WesleyDavid Still wouldn't eat it
@ewwhite Leter 'S' versus letter 'L'
I'm not helping, am I. =(
@ewwhite Isn't it single/multimode?
@WesleyDavid actually that's close to what our storage guy used to tell the datacenter cable monkeys: "SC" Small Connector, LC "Large Connector"
Or rather, one's designed for single mode over long distance and one's designed for multimode over long(ish) distance?
22:18
Debating on whether the two network handoffs from my new data center should be copper or fiber.
It's been a long time since I had to bother with fibre connectors
I have a pile of SX modules here...
@MarkHenderson Yes-ish - SX is multimode and LX I think can be single or multi
but the Data center wants LX
@ewwhite Given the option, do they offer 10Gbit over both?
22:19
@ewwhite Interesting. From memory our DC only offered SX
@voretaq7 Gigabit uplinks.
@ewwhite fuck it, copper.
Jiggerbutts.
@voretaq7 Any difference?
I've taken to pronouncing "Gigabyte" as "Jiggerbutt." Because.
22:20
IMHO it's not worth the expense for fiber unless you've got at least the option of multi-Gb expansion in the future
@ewwhite Yeah - everything has gigabit copper and it costs $0. Fiber would cost money :-P
I have the SFP's... just wrong type.
and it frees up a couple of ports...
that I could use elsewhere... for systems co-located by @JoelESalas
@ewwhite that's a minor advantage I guess but you'd have to change out the modules i think, and those are spendy
@WesleyDavid Do you approve? cl.ly/image/0i0p0K2Q0r2q
@mossy Polydactyl cats FTW
haha
22:34
@LucasKauffman Enable "offsite forms"
blogs.technet.com points and laughs
Oooo, Microsoft. You wiley coyote.
The entire domain was throwing 503 errors.
Works now.
They must have rebooted their cloud.
@WesleyDavid stalling for me.
ah, there it goes...
nope - still stalled. Just loaded the <title>
@ewwhite Kewl.
Oh, web hosting. Right up my alley.
Hey guys, #semove hashtag on twitter for live updates of the SE datacenter move.
2
Where are you moving the datacenter to this time?
@MichaelHampton The basement.
@PeterGrace Are you guys doing custom-sized power cables?
(if yes: "Fucking OCD sysadmin bastards!" ; if no: "Lazy sacks of shit!" :-P )
22:51
@PeterGrace fix my chat screen!
@ewwhite REBOOT!
In Chrome, my screen jiggles when I'm in the Comms Room.
That is all.
Hot.
@ewwhite Jiggle Jiggle.

« first day (856 days earlier)      last day (4411 days later) »