« first day (144 days earlier)      last day (4820 days later) » 

4:00 PM
it's optional
 
@Ben still got the same error
 
phew - glad that ordeal is over - how is everyone?
 
@barraponto these command sshould be safe
only the first 3 are global
 
hey there @Chopper3
 
4:02 PM
zend framework is such a headache .. first headache is while learning ... second headache is when developing ... and the last headache is when uploading.
 
Ben
@explorex Wow.. hmm. You'll need to dig up your server config, and see what's going on in there. You may have a 'deny' set for that directory or something.
 
@Ben
which means
 
@explorex is AllowOverride All set in the main httpd.conf?
 
@Josh i think i don't Webhost Manager can access the whole server please check my question again i am going to upload snapshot of my web host manager
 
@explorex is this your server, or a shared hosting account?
 
4:10 PM
this following line, meant to be used on cron level: /usr/bin/rsync -aqz --delete rsync://drupal.server.example.com/example-cache /var/www/example.mirror/cache/normal/example.com/
 
Is anyone else seeing a strange UI bug between tools and Chat in the SF bar?
 
shouldn't it read rsync://myserver/cache (since settings are called [cache] on rsync.conf)
 
(Tools is the 10K tools, so not all of you will have it)
 
woo, got my new 3750x stack
 
@Josh some one created some domain and gave me all details not my server
and please have a look at updates
 
4:13 PM
@barraponto hmmmm
 
@sysadmin1138 mine looks normal to me
 
@explorex yeah, you do not have accss to modify httpd.conf
Talk to your server admin about this issue
@sysadmin1138 I don't see that but I don't have 10k
 
@Josh thank you i am going to talk to him and thank you very much
 
@sysadmin1138 don't see anything, but i don't have 10k either
 
Thank you all. Maybe I just need to hit Refresh harder.
 
4:16 PM
@barraponto I think you'recorrect
 
Ben
@explorex you must have a "deny from all" set somewhere.. did you post the entire .htaccess file? are there other .htaccess files above it in the directory? can you load your index.php file directly?
 
@Josh this is only the one i have
 
@Joshcan you point me to a good read on rsync? better than manpage
 
but if i put different html file then i can get the url
@Jo
@Josh besides there is another file phpinfo.php and test.php and i am not being able to access that file too
 
@barraponto manpage is my favorite.... um, does anyone else know of a good rsync tutrial?
@barraponto I rsync my servers using rsync /path/to/files destinationHostname::rsyncConfigOption where rsyncConfigOption is the entry in /etc/rsync.conf on the destination
@explorex yeah something with your rewriterules is not working
 
4:24 PM
may be the rewrite_mod is not installed
 
@Josh will these options be available when cron runs the command?
 
@Josh i deleted the whole directory and i have phpinfo on index page, can i know if mod_rewrite is enabled from phpinfo?
 
@barraponto they should be, yes
@explorex ummmm, I am not sure
One moment please
 
usually not even $PATH is available at cron runs...
 
@Josh link is here websmartint.com
 
4:31 PM
@barraponto that shouldn't be an issue as long as you say /usr/bin/rsync instead of just rsync
 
ok thanks
 
@explorex create a page 'test.html' and put something in it, like <h1>mod rewrite works!</h1 and then place the following into .htaccess`:
RewriteRule .* testPage.html
 
okay
 
where testPage.html is the name of your HTML page
then, go to any URL and see if you get your test HTML
 
4:40 PM
@explorex ok, did you make the .htaccess?
RewriteRule .* testPage.html
 
yes
and i have testPage.html
file
 
looks like mod_rewrite is not available
Can anyone else confirm?
 
okay i will try at my localserver
 
Why is the process of bootstrapping major upgrades always so painful?
 
@voretaq7 They wouldn't be called major upgrades otherwise. Major == bigger than what you can take on a regular day
:)
 
4:51 PM
heh
 
I specially hate upgrading CentOS, if only it was as easy as Ubuntu...
 
dude, ubuntu upgrades SUCK. apt is a pile of festering corpses.
(the aforementioned major upgrade is Ubuntu 9.x to 10.x)
 
apt is cool when you start to know it, but the learning curve is a bit to steep for my to have any simpathy towards it
@voretaq7 Ubuntu 9.x to 10.x should be okay to do though
 
@voretaq7 i would posit ALL OS upgrades suck
 
@lynxman Breaks (consistently) on the *SECOND* reboot
 
4:53 PM
I've done 8.04 to 10.04 couple of times
@voretaq7 What errors are you getting?
 
I've had to go around the normal update process and engineer my own (which is better in my case, because it gives me more positive control than apt and doesn't break on the second reboot)
 
@voretaq7 If you can send me the errors you were getting I can make sure to try and make a bug request out of it
 
the crux of it seems to be some kind of gnome-related insanity -- I wasn't interested enough to really troubleshoot b/c it's got to be something tied up with our customizations
 
ah, the much feared gnome dependency labyrinth
 
I can't really blame ubuntu because updating stock 9.x to 10.x worked fine - they're just the focus of my rage because things shouldn't break on the *SECOND* reboot :-)
@Zypher *BSD upgrades suck a little less
 
4:56 PM
rofl
 
@Josh i found out that rewrite rure is working fine
 
*BSD == make world, make install world, reboot
 
check it out i has just missed rewrite engie on on htaccess file
 
@lynxman in my case add "run mkpatch ; run cluster_upgrade" to the end of that :)
 
@voretaq7 nice
@voteraq7 all being said, if I can give you a hand in anything Ubuntu related let me know
 
4:58 PM
@Josh anyway thanks i will give my last shot if not then i am going to quit
 
I hang around #ubuntu-server on Freenode
 
eh, not as nice as it should be - we're using radmind, which is pretty good. It's also what's handling the ubuntu upgrades once I get the bootstrapping bit done
We need a service where I can rent a junior sysadmin for like 2 weeks at a time -- I've done all the hard (and interesting) stuff, and now it's just the annoying little things.
 
@voretaq7 so just boring grunting job left to be done
 
@lynxman yup - tweaking permissions in the bootstrap packages, adjusting the template system to eliminate a few errors, and then I can unleash this on unsuspecting users :-)
oh left out a bit on those BSD upgrade steps: "portupgrade -af" :-)
 
@voretaq7 hehe, they won't know what happened, they'll just get a stable upgraded platform
 
5:08 PM
@Josh i nailed where i got error
 
@lynxman That's the plan. One day I'd like to get our field units off of Linux/Ubuntu and onto a BSD platform like the servers.
the barrier to transition is pretty high though: Linux has the desktop/GUI thing pretty well sorted
 
@voretaq7 true, specially Ubuntu
 
Today's News
TechNet Wiki: Windows Azure Survival Guide

The Windows Azure Survival Guide is a great resource for the Windows Azure community of IT Pros. Containing links to videos, blogs, forums and a host of other useful links; it can help you navigate the wealth of Azure information available.
Alan Le Marquand, MSFT
http://social.technet.microsoft.com/wiki/contents/articles/windows-azure-survival-guide.aspx?ocid=otc-f-us-jtc-DPU-TNHEADLINE
 
0
Q: rewrite rule for .htaccess not working for php file

explorexWell, to my previous question i found out the RewriteEngine On RewriteRule .* index.php was not working but RewriteEngine On RewriteRule .* index.html was working, what could be the possible reason

 
@lynxman yeah I give them a lot of credit for that. There are a few pain points, but that's mostly because we're using these machines in a funky way (field units at hospitals, so auto-login, remote assistance, etc. -- it's kinda kiosk-like actually when you get right down to it)
 
5:15 PM
hi anyone viewed my question?
 
@voretaq7 Maybe you should get some support on those Ubuntu installs
so you can take that off your head
 
eh, they're not hard to support once they're in the field (and support in the US anyway needs to stay in-house due to the HIPAA hippo in the room). It's the deep customization we do that's annoying.
Hopefully we'll be able to justify a junior admin to handle it this year though -- it'd be nice if someone were keeping up with Ubuntu development and working on the OS full-time :)
 
@voretaq7 That's what I'm talking about support, so you can get that delivered to you
or someone to throw this problems at
 
I'd love that - problem is we can't justify the staff right now (as annoying as it is there's not enough work to keep someone busy full-time. I even have time to troll for rep here while I wait for builds to finish :-)
 
@voretaq7 Yeah, same here (launchpad why are you so slooow)
 
5:29 PM
mornin
 
hey @SpacemanSpiff morning
 
so... today's problem... figure out some clever way to route two like subnets.. without using NAT
 
'day
 
both will have a common destination... i don't know this is possible
 
@Iain dropped the g?
 
5:34 PM
@lynxman saves having to open a Fosters
which is always a bad thing
 
@Iain hmm good point
@Iain specially in the UK where Fosters tastes like racoon piss
oh gosh, not again
 
Nearly all Australian beer (especially in Aus) is cold, fizzy, piss - the exception is Little Creatures IPA
 
0
Q: apache settings as backend behind nginx

Mike Williamsi have a 2gb ram / quad core server and im running nginx as a reverse proxy but what should my apache settings be for best optimization to handle ALOT of requests and how can i test it under load? current: Timeout 300 TraceEnable Off ServerSignature Off ServerTokens ProductOnly FileETag All Sta...

please give me all the details I need for my completely undisclosed web app
you should know! you smart engineer!
*pouts
@Iain *IPA is normally awesome, but I'm just biased like that
 
IPA is my favorite
 
any kind of IPA
 
5:39 PM
@Iain I live for a yummy brown
 
I still have dreams about Portland IPA selection, that was life
 
@packs a what ?
 
@Iain Brown ale. Newcastle is a well known example.
 
Wae'aye man
I used to drink it but I prefer them lighter these days
 
0
Q: How to set up a file server in a restricted corporate environment

Emilio M BumacharI work in a big corporation, and the disk space my team gets in the corporate file server is so low, I am considering turning my work PC into a file server. I ask this community for links to tutorials, software suggestions, and advice in general about how to set it up. My machine is an Intel Co...

Guerrilla networking. Gotta love it.
 
5:43 PM
also known as "please let me add more holes in the corporate network to make the security sysadmin life unbearable"
2
 
@Iain That reads like some kind of Northern abuse of the language....
@sysadmin1138 At least he upgraded to SP3....
 
Geordie colloquialism in reference to Newcastle brown
 
@packs Small favors make all the difference.
 
I love it! "A lot of security software I don't even know about is and must remain installed."
 
@sysadmin1138 the fun part is when some department manager comes back to you and says "yeah this NAS hard drive we bought a few years ago that has a few terabytes of data on it that we need isn't working, can you help"
 
5:48 PM
@SpacemanSpiff In my experience, more like a few months.
 
Thank you, Evan.
 
You turn the device on, and realize that more than one of the heads on the disks are clicky-clacking....
 
Followed by something akin to, 'Just support it gorramit! This is mission critical to our department and cannot be removed!'
 
running some random software RAID that puked its brains out
 
Election Ends in 3 hours, if you han't voted Please do! serverfault.com/election
 
5:50 PM
Is it worth a short-duration system-banner?
 
As it turned out, his employees were too lazy to upload their work to our NetApp at the end of the day
so they'd pass this drive around :O
 
hmm maybe
 
@SpacemanSpiff :0
 
we should have a close route for This is a really bad idea ...
 
5:54 PM
@Iain flag as "you crazy!"
 
Dec 15 '10 at 18:27, by Robert Moir
Maybe that should be a new reason for voting to close?
*Closed: This is a dumb idea and you really shouldn't even be asking it*
@Iain It's in the star list :)
 
Hello rep cap, how are you... heh heh.. I love questions like that "rogue fileserver" question.
 
off to watch some football, cya later guys
 
@EvanAnderson Well stated, by the way. I may have to shamelessly rip off your answer the next time something similar comes across my desk
 
@packs: Thanks! That's what I'm here for... rip away.
 
6:00 PM
Normally, I fall back to the "You are free to do this as long as you want, so long as you realize switchports will seemingly disable themselves along the way." Effective, but not very diplomatic :)
 
he he he he he he he...
 
@packs I'm personally a fan of "ARE YOU OUT OF WHAT IS LEFT OF YOUR OBVIOUSLY VERY LIMITED MIND?", but I've learned not to say that (just think it very loudly)
 
@voretaq7 That kind of direct confrontation tends not to go over well in Higher Ed :)
 
@packs the minds tend to be more limited there too :-)
(also doesn't go over too well in the corporate world either)
 
As every public employee in the room starts inching slowly away from the source of that statement...
 
6:15 PM
Time for coffee!
 
@sysadmin1138 :)
 
@EvanAnderson damn annoying isn't it :(
 
Did I really see someone want to start up lá networking resistánce on their workstation in the list of questions?
 
Yes you did.
 
dear lord
 
6:23 PM
what? where?
 
mr XP is my file-server
woohoo I am being trained in "customer service" tomorrow
 
SysAdmin1138 Expounds
Rogue file-servers
SysAdmin1138
http://feedproxy.google.com/~r/Sysadmin1138/~3/Ux9Bq7o-KoU/rogue-file-servers.shtml
 
@RobertMoir Congratulations on the promotion
 
now that's a fast turnaround
promotion?
 
Customer Service... never mind. It was a rather weak attempt at smarmy wit.
 
6:27 PM
lol
i see
 
@RobertMoir Mr XP-Is-My-File-Server is a good candidate for being the canonical question for 'how do I bypass my stick-in-the-mud IT people'.
 
Yeah
 
It may not be the IT people, it could be some overly strict legal/corp policy or something else that is very inflexible.
 
yep... we have to "enforce" web bans with websense and get the blame for other peoples rules at my place
 
So the IT defaults to setting up storage with HA, huge backup retention, when users are looking for scratch storage. OTOH, sometimes users aren't very good at expressing their needs, and what starts off as scratch storages frequently becomes mission critical...
 
6:31 PM
yes indeed
 
Backing up an 80GB server through a 2MB pipe is like attempting to drink a 72oz MegaGulp through a cocktail-straw.
3
@Zoredache That was our problem in a nutshell. We had one primary storage tier, and it'd be the top tier any place else that had multiple tiers.
AHAH! The mystery of the UI bug is solved. It's the new Pending Edits notifier.
 
uh huh
 
@Zypher: So, ever decided on the ASA/3750/routing setup? I figured out mine, toying with 2x3750x (stack) and 4 2960S right now
 
I suppose only 10k+ users receive the notification?
 
@packs That's my guess. I'm not certain.
 
Ben
6:34 PM
That is just silly if it's 10k+
 
3
A: Edit a question (from other user) and get +2 rep ?

Ivo FlipseThis just got implemented: So nope, it's a feature not a bug!

Yep.
 
I cant understand why one needs 10k rep to see suggested edits..
 
Ben
Me neither, since a 2k user can approve them anyway
 
i dunno
 
@pauska That's the thing. Approving edits is much lower, but you just can't see them unless you happen to open that post?
 
6:36 PM
it does seem a little odd
 
Ben
It's the same with close votes - we can cast them, but not see where they have been cast so we can (possibly) help it along
 
@Chopper3: Annoying? Oh-- the rep cap. Yeah. I haven't seen it in awhile, though. It's like an old friend.
 
I kissed it yesterday. Somehow didn't increment my 'rep-cap reached' count though.
 
@Zoredache: There is so much truth to your statement "...and what starts off as scratch storages frequently becomes mission critical...".
 
@Zoredache if you ask enough people everything with a plug on the end will be mission critical
 
6:43 PM
I haven't done a count of my rep cap days in awhile. That "Legendary" badge is still 65 rep cap days away... >sigh<
 
I'm 22% of the way to Epic!
 
I've done about 28 rep cap days , so past the half way point to epic but you should see the number of 180-200 days - SO many
mind you, with new mods coming very soon I might be able to answer more perhaps?
 
Dan
I have 1 rep cap day!
 
fox news - putting egypt in the wrong continent!
 
@Chopper3 They're doing a fairly good job of balancing the landmass.
 
6:51 PM
never let accuracy get in the way of news
 
@Chopper3 I'd like to see a few more sources for this major screw-up. I'd also like to see Geography taught in US classrooms again.
 
what? it's not now?
 
7:07 PM
@Chopper3 not as a discrete subject, no. Why do you think most americans have such a piss-poor concept of where countries are?
 
but..but...but...* shakes head *
 
My personal favorite idiot geography quote: "Africa is such a poor country". When I confronted them with the fact that it's a continent with several countries in it their reply was something to the effect of "No, it's a big country in Europe somewhere".
(I was sputtering for a month about how our educational system is a dismal failure)
 
is the location of Egypt really a geography question? Isn't it more of a political question?
 
I'm not one to talk about geography though, I probably can't place all 50 states on a map accurately. But I can sure as hell give you the correct continent & region :-)
 
@voretaq7 I would hope that would be an isolated incident rather than an indictment of our, admittedly neglected, education system.
 
7:10 PM
@Zoredache What with the Nile delta being clearly visible in the photograph I think it would be easy to place Egypt correctly - I consider that to be geography :-)
 
@Zoredache I'd say it's very much a geography question, given that's Iraq on the photo - you know, that place where we all started shit a while back
@voretaq7 a while back I figured out I'd been to (however briefly) 42 US states - quite pleased with that, might try to finish the lot in the next few years, just for the hell of it
 
@Chopper3 I've passed through all 48 contiguous states if you count flying over them. I've been on the ground in like 30 I think
also where "we" = "The USA" then "Place where we all started shit a while back" is pretty much every country, isn't it? We like to meddle. We're the world's Mommy Dearest.
 
actually I think most of the world's troubles can be traced back to English meddling between 1830 and 1940 - we really messed with the world - loads of it is 'our' fault, you guys just grabbed the controls and have taken over
 
@Chopper3 well we learned how to be a mother from our mother country :-)
 
haha, seriously though, all the trouble between india and pakistan, between north and south sudan, so many other place all started when us brits drew up borders and segregated places
 
7:26 PM
@Chopper3 I wish I could have been a fly on the wall in that map room. "Just draw a fiddly line here, and here...done! New country. Next!"
 
all done for astonishingly selfish reasons, no change there then! :)
how long until the results now? 30-something minutes?
 
Ben
I dunno, are they announced at 9 or do they have to be manually calculated and posted on the blog (for example)?
 
@Chopper3 It has gone on long enough that I have lost interest
 
:)
 
Ben
I have to admit I've tried several times to understand how the voting system works, but my brain can't cope with it :/
 
7:28 PM
I just want a bit of help (although it's been a bit better recently)
 
@Ben Right. Instant run-off I understand, that one's easy.
 
Ben
I'm sorry?
 
As near as I could tell from the voting method that they're using, it is trying to accomplish the same basic thing that instant run-off does. Only using grad school level stats instead of addition.
 
Ben
Righty. The Wikipedia page they link to with an explanation doesn't particularly explain (IMO)
 
@Ben That's the problem...I think it does....
 
7:43 PM
I meant to do a blog post about how it worked, but never got around to it.
 
educate us while we wait then @sysadmin1138 as I don't understand it
 
I last looked at it close to two weeks ago, so this'll be imprecise.
In this kind of election, a victory threshold is chosen. That threshold is set by the people running the election, but could be 40% in our case (two position, minus a fudge). The first pass of the election counts everyone's 1st Choice votes. If one candidate exceeds the threshold, a random sample of 1st choice votes are selected from the pool who voted for that person and their 2nd choice votes are promoted to 1st choice, and allocated among the remaining candidates.
 
oh @Chopper3 you see this blog post:
Posted by Jeff Atwood on February 2nd, 2011

First, a quick update on the Stack Exchange moderator election schedule.

Ending today, so get those votes in!

Ending soon:

Just started:

Coming soon, in this order, at 3 per week:

(as always check out Yi Jiang’s awesome Stack Exchange Moderator Candidate Statistics page for insanely detailed election stats on any of the above.)

As you can see, we have a lot of community moderators! That’s why we’ve been working so hard lately on improving our moderator tooling, to make sure our new class of incoming moderators have as pleasant, polished, and smooth an experience as possible. …

 
Alternately, if NO CANDIDATE crossed the threshold on the first pass, the candidate with the LOWEST 1st-choice proportion is eliminated and all of those 2nd choice votes get promoted to 1st-choice and allocated among the remaining.
 
scratches head, drinks more coke
 
7:52 PM
Wash-rinse-repeat, until you get your two 40% moderators. That's the simple version.
 
With the goal being that the electorate are so confused by the process they just assume the winners are for the best?
 
Thanks - clearer than it was before
 
I believe the goal is to have two moderators that a majority of people actually voted for in some way. It may have been 3rd choices all around, but they DID vote for them.
 
it'll be alright on the night
 
heh. Apparently this is instant runoff
Just by a different name
 
7:55 PM
I'm trying to figure out how it differs, but I can't crack it.
 
Well, according to this part of the wiki page, and assuming your description is accurate, I can't see it either
Instant runoff voting (IRV) or the alternative vote (AV) is a voting system that elects one winner by ranked choice voting. Instant runoff voting is designed to more accurately represent voter preferences than the first-past-the-post voting system. Procedure As a form of preferential voting, voters mark candidates in order of preference with the numbers 1, 2, 3, etc. When IRV is compared with First-past-the-post in single-person elections it eliminates vote splitting, reducing concerns about tactical voting and strategic nomination (Arrow's Impossibility Theorem). The winner is decided...
 
Has anyone said when the results will be known ?
 
not to me
 
All i know is that I'll be obsessively refreshing several places in about 55 minutes.
2
 
hahaha
 
8:06 PM
I grew up in a family where election-watching was a spectator sport. I've been here before.
 
uh huh
 
I grew up in a weird family.
 
I have a friend who lives in Eire and she says our elections are boring because the results are known next day
 
I can't believe the question about how to verify your backups is still going.
 
where is that ?
 
Ben
8:08 PM
Whaat
 
3
Q: How do you verify a restore?

NicWhat tool(s) would you use to verify that a restored file structure is whole and complete? My environment is a Windows Server 2008 file server. (We use tape for backup, but that is inconsequential.) I am specifically looking for a tool that will: Record the names of all files and folders below...

 
Ben
Restore. <Obligatory filler space>
Easy as that ^^
 
I thought it was fairly obvious, best way to check a backup is to restore it. Anything else is a compromise.
In fact, to hell with it, I think I'll put that as an answer.
 
@Iain All the fun is in watching the hour by hour updates!
 
@Iain Downloadable results are avaialble immediatly so you can run them yourselves if you want ... gonna be a little while after that for the official announcement after we verify the data
 
8:10 PM
@sysadmin1138 so she tells me
@Zypher I'll pass
 
I sense some time with a spread-sheet about 1pm my time.
 
@RobertMoir Mark has it right on that -- question: Rebuild your environment using nothing but clean OS media and your backups. If you can't do it you failed the test.
 
Yeah, we had to do a restore of a server the other day - running windows DHCP and certificate server roles. There was some oddities from that combination of roles but it worked. Now the oddities are known they're in the docs
 
@RobertMoir Restore procedures? That's crazy talk
 
We routinely rebuild our dev environment from bare metal every 6 months (thank you VMWare) and restore backups from the production environment to repopulate its databases.
 
8:17 PM
Huh. The super-computer that's going to be on Jeopardy soon is running SLES, not RHEL.
 
@Voretaq7 that's definately the way to do it. Also, I <3 VMWare
 
@sysadmin1138 Not surprising, IBM was very much a SLES shop for a while
 
@sysadmin1138 So it'll speak with a weird accent and die halfway through due to package inconsistencies?
 
@RobertMoir It's a fantastic restore test Only problem is it's kinda time-consuming (takes a full day)
but knowing we could rebuild out entire production environment in 8-12 hours in the event of a catastrophic failure is very comforting :-)
 
I bet it does, but better to spend a day under relatively little pressure practicing it than having to do it blind at 3am with the big boss screaming and crying
3
 
8:21 PM
Taking out one day once or twice a year is a small price to pay for having a well tested, and known working, DR plan.
2
 
I'm sad to say my boss won't allow us to do the proper restore test thing so I'm not practicing what I preach at the moment. But the team know the score and we try and do our own rehersals and make the most of any chance to practice a restore process properly.
 
@packs At some point I'll actually do a full drill (right now we pre-stage the restores from production so we don't have to sync the entire off-site Virtual Tape Library) - we expect that will add another 6 hours because you have to sync a few gigs of tape files before you can use 'em
 
@voretaq7 Still, you're already head and shoulders above just about every other company I've had the pleasure of seeing.
A lot of places will go as far as building out a plan, but are terrified of testing it.
 
but we've proven the three important things: You can restore the build/provisioning server (so you can rebuild the rest of the environment), the LDAP server (accounts/passwords) and all of our databases. Everything else is gravy.
when I came here they had a system built by their previous admin. I got them to test it and it couldn't come within a thousand miles of a functional environment when they restored it bare-metal. He apparently never tested his restores.
 
ouch
 
8:26 PM
that was a very tense 6 months of designing and implementing something that actually worked :)
 
I'm reasonably confident with what we do - I've tested restores of most things over time, just not all at once, which bothers me but I can't push that through sadly. I'm envious of voretaq7's structured approach, I really am.
my geek spider senses are tingling. It's been a while since I upgraded something in my home rig.
 
Arg. That 80GB backup over the 2MB link looks like it'll take over 90 hours to finish. Once it's done, though, it's all incrementals.
 
I generally don't do to much full restore testing, but I always do have at least two backup methods and sometimes 3
 
@sysadmin1138 its taking ages to do the backups to crashplan for my home gear but when its done its done
 
@Zypher, @KyleBrandt, so how long until the stackexchange network supports IPv6? Apparently they want you to enable IPSEC as well. meta.stackoverflow.com/questions/69171/…
 
8:34 PM
ehh IPv6 is on the "when we have nothing better to do" list
 
@RobertMoir In this case, in a disaster it'll be faster for us to drive down there, haul the gear back up here, restore, then drive it back down.
 
and yea ... about that IPSEC stuff
 
@AntoniusBloch you don't have to be obsessively restoring stuff all the time but its just about having confidence in your own backup routine. At the end of the day its we're all out of our respective jobs if we can't get the data back, right?
 
For rsync backups I have a little script that writes a little random file on the backup target and then compares it with the backed up files.
it is a little redundant
 
OMG FIRE FIRE WILL FIX NOW!
 
8:35 PM
OMG Ponies
 
Sorry guys, have to take everything down for maintenance right NOW!
 
tense
 
eyes suspiciously
 
j/k??
 
Just in time for the election to end. I have faith in the system now :)
 
8:36 PM
he has to see some guy called chad, who's hanging about the voting papers
 
Ya, responding to at @Zoredache
Apparently I need to spend the $20 dollars needed to implement IPv6 and SSL and IPSEC etc -- that is all there is to it
 
let's say you have an v6 address
can I reach it right now?
 
just give us HTTPS, all you have to do is spend ~$200 on a cert from Godaddy, right?
 
@KyleBrandt $20?
 
or does that depend on my ISP?
 
8:38 PM
Or did SOMEBODY get bit by the security-bug?
 
we're looking at IPv6 because the UK academic ISP, which we're a part of, is considering it. It's implemented at some universities already
 
@AntoniusBloch, possibly, you would have to setup a tunnel.
 
@RobertMoir Both our upstreams support it. We need to finish changing out all of our edge switches before we can even consider announcing anything on the network.
 
@Zoredache: I think I have an old pentium 3 around for the SSL, will that handle the load?
 
Something like Teredo allows you to pass IPv6 across the IPv4 network.
 
8:39 PM
just use an old vic 20 and be done with it
@sysadmin1138 yeah the support is there if you need it, but there's always something more important to look at first
 
@KyleBrandt, probably not. But I am pretty sure Jeff mentioned in his recent blog, about the updated search engine, that your web tier is nearly idle all the time...
 
@Zoredache turning on SSL can fix that :-D
 
heh
 
@Zoredache we fixed that problem
 
@RobertMoir our approach isn't as structured as I'd like -- I'm actually backing up more stuff than we need to technically, and we have no deduplication in place at the moment :(
 
8:41 PM
I think badges should give points
 
@voretaq7 we're using commvault galaxy so we have de-duplication coming as standard in the next version of that I think
 
@RobertMoir I'm not sure if/when Bacula will get deduplication. It's made a bit harder by the fact that we have to encrypt everything before it can go to storage too...
 
ahh that's not going to sit well together then
 
@RobertMoir luckily we've eliminated the bulk of the OS by backing up the build/provisioning server so instead of backing up 12G on every box we only grab it once :)
 
Nothing like introducing whacking great amounts of entropy to really kill a de-dup system.
 
8:44 PM
oh yeah
 
@sysadmin1138 I suppose I could just use the Debian version of OpenSSL for the encryption. That one compresses beautifully!
 
XOR is nice that way.
 
Some things you never live down. Removing entropy-gathering from your crypto code is one of 'em.
 
:-)
 
Rule #2 of systems work: "If you don't understand it, don't mess with it".
2
 
8:46 PM
rule #1 of where I work: "If you don't understand it, leave it on Robert's desk with a note saying the deadline was yesterday"
 
@RobertMoir I've had that job-position.
 
I think its just about every IT job, in fairness
 
I have that job position :)
but I prefer that to messing about with systems (and of course not backing up the files you're changing, or recording what you're doing), breaking them horribly, then coming to me and saying "The production database server is down, any idea why?"
 
uh huh
 
I thought it was "If you don't understand it, fiddle with it until it breaks, fix it, and then you will probably have a decent understanding"
 
8:48 PM
@KyleBrandt True, dat.
 
And I like saying "Because you deleted all those pesky .log files. I love auditing, don't you?"
 
@KyleBrandt That's rule #3 ("Since you ignored rule #2 and broke it, now fix it. Rule #2 will no longer apply after this")
Rule #1 is "Stay the hell out of my production environment!" -- at least when I'm running the production environment :)
I'm looking for a good rule #4
 
its not always us who have to deal with 'em!
rule #4: If you ask the sysadmin why you can't log on during a powercut, you will receive a snarky answer.
 
Things I Love: Patching the dev environment, then rebooting 15 virtual machines at the same time.
 
i bet the devs love that
 
8:52 PM
@RobertMoir By "snarky answer" you mean "slammed head first into the open power panel"?
@RobertMoir They requested these changes :-)
 
only in my dreams
its all gone quiet
 
@RobertMoir sorry. it's a really crazy day for me over here :-p
 
ahh hope it calms down soon Josh
 
Almost time to start hitting refresh.
 
hits refresh
 

« first day (144 days earlier)      last day (4820 days later) »