|
larchesdanrew posted:"My computer is really slow and Premiere crashes when I try to open it. " See, hard drives do break when they get too full. 
|
#
?
Sep 24, 2015 16:12
|
|
|
|
| # ? May 17, 2024 19:33 |
|
|
103%. That's completely bonklers.
|
|
#
?
Sep 24, 2015 16:13
|
|
|
Truga posted:103%. That's completely bonklers. I'm a good math man 
|
|
#
?
Sep 24, 2015 16:18
|
|
|
Malachite_Dragon posted:Don't you work for a school system? Who the hell DDoS's schools? Students. Back in 2000-ish roughly half of our school's IT Department was volunteer student workers and sometimes they'd bring the whole network down for a period with an "attack" if they wanted an extra study hall. They also hosed around with the PA system from time to time. Having "The Wall" blaring over the speakers as students left on the last day of school was pretty fantastic.
|
|
#
?
Sep 24, 2015 16:27
|
|
|
Sefal posted:Reading this thread somehow makes we wanna keep learning stuff. U guys inspire me. In my experience, the best way to learn is to break something in production by being an idiot, and fix it before alerts start going off. That's how i started in my career, by being a moron, misplacing a "}" in a dns config and freaking out when the service stopped and didn't restart. Thats also why I post about the stupid poo poo I do (and others, but primarily me) so others can learn from my dumbness.
|
|
#
?
Sep 24, 2015 16:28
|
|
|
nitrogen posted:In my experience, the best way to learn is to break something in production by being an idiot, and fix it before alerts start going off. I once brought a production Xsan down by propagating permissions from the root volume, it was my second week on the job. I learned that day all about ACLs and POSIX permissions on a Xsan and how to propagate them properly in an Xsan environment. Alternatively a good (and safe) way to learn is to build a lab environment and then break the crap out of it.
|
|
#
?
Sep 24, 2015 16:37
|
|
|
luminalflux posted:AP chat: We're an open plan office in downtown SF, old brick building (20+ stories), lots of people on wifi, everyone streaming twitch all the time ( I have a site with 128 Arubas and like you said if you take time to get it sorted out they seem reliable.
|
|
#
?
Sep 24, 2015 16:42
|
|
).
|
We're phasing out our Aruba gear. It was nice when we got it, but performance has been getting worse and worse as time went on, and support has never been useful enough to justify the enormous cost compared to their competitors. As far as I know we're the only one with negative feedback for Aruba though.
|
|
#
?
Sep 24, 2015 16:51
|
|
|
As a note to the people who have had issues with UniFi APs. Several years back, one of UniFi's manufacturers (in china, of course) broke contract with UBNT, kept the equipment designs, and started producing and selling the gear by themselves, including the UBNT logo. They were using substandard parts, and the copies from this manufacturer were breaking left and right. Since they LOOKED real, resellers assumed they were legit. I believe UBNT finally forced them to stop, but the designs 'leaked' and a few other places started making copies in china and russia. So, long story short, make sure your poo poo isn't counterfeit. If it isn't from a vendor listed on their website, there is no guarantee it's real.
|
|
#
?
Sep 24, 2015 17:04
|
|
|
pr0digal posted:I once brought a production Xsan down by propagating permissions from the root volume, it was my second week on the job. I brought down an entire building of 1,000 users by changing the near end of a trunk link first. Fortunately, the building was only 200 yards away from the main building and I managed to dash over and get it sorted in 10 minutes.
|
|
#
?
Sep 24, 2015 17:05
|
|
|
Installing Yosemite on an older (but not much older, only 2007!) iMac with a 1440p screen. The user previously had it set to 720p, presumably so that he could see things better. Does OS X have a global "UI size" thing like Windows does? I can't seem to find it and Google just suggests tuning things per-application.
|
|
|
#
?
Sep 24, 2015 17:21
|
|
|
Segmentation Fault posted:Installing Yosemite on an older (but not much older, only 2007!) iMac with a 1440p screen. The user previously had it set to 720p, presumably so that he could see things better. Does OS X have a global "UI size" thing like Windows does? I can't seem to find it and Google just suggests tuning things per-application. Try doing this: http://cocoamanifest.net/articles/2013/01/turn-on-hidpi-retina-mode-on-an-ordinary-mac.html 27-inch iMac should have 1280x720 hidpi mode, which will be pretty cramped but sharp instead of blurry. That's really the only way to do it on OS X, I think.
|
|
#
?
Sep 24, 2015 17:34
|
|
|
Oh stupid poo poo we've done in production? Accidentally dropped the Users table in a production database because I thought I was connected to the database in test. (Don't have your production database server accessible from the test environment. Firewall at the least, airgap if possible. Don't make the test database have the same name as the production database.) I've also applied a switch configuration which brought down the entire access network and locked me out of the switch. I had to interrupt a board meeting (because access to the network cabinet was through our largest meeting room ) to plug a serial cable into the switch.
|
|
#
?
Sep 24, 2015 17:37
|
|
carry on then posted:Try doing this: http://cocoamanifest.net/articles/2013/01/turn-on-hidpi-retina-mode-on-an-ordinary-mac.html He was running 720p anyway so the screen real estate is essentially the same, everything just looks nice and sharp now. Thanks!
|
|
|
#
?
Sep 24, 2015 17:59
|
|
|
Crowley posted:Our ISP in on the case. I just opened a ticket saying something like "this line is unstable. WTF guys?" and they took it from there. My only problem right now is being allowed to give them some money for their DDoS guard. One of our ISPs actually dropped us during a DDoS because we were affecting their infrastructure. Management still won't let us tell them to gently caress off because they pour advertising money our way.
|
|
#
?
Sep 24, 2015 19:46
|
|
|
larchesdanrew posted:"My computer is really slow and Premiere crashes when I try to open it. " Regarding stupid mistakes, sometimes I don't notice if computers are on before doing hardware stuff. The other day I pulled a DIMM from a running PC. Windows didn't like that. The worst was several years back when replacing a bad IDE CD drive. Got a nice spark when I plugged the new drive in. Never heard back from the user so I guess I didn't fry it.
|
|
#
?
Sep 24, 2015 20:21
|
|
|
Knormal posted:Sounds like somebody needs a Buffalo ASAP! WTF? Hasn't SATA been standard for the last decade?
|
|
#
?
Sep 24, 2015 20:41
|
|
|
ilkhan posted:I...D....E....? Several years can include 10 of them.
|
|
#
?
Sep 24, 2015 20:50
|
|
|
deimos posted:One of our ISPs actually dropped us during a DDoS because we were affecting their infrastructure. Management still won't let us tell them to gently caress off because they pour advertising money our way. We're a public institution on the national contract. I don't think the ISP can drop us. On the flip side we can't drop them either, and they're not exactly fast on getting that DDoS blocker up and running.  In all fairness their lines are rock solid once they get installed, and super cheaply priced too.
|
|
#
?
Sep 24, 2015 21:39
|
|
|
Collateral Damage posted:Oh stupid poo poo we've done in production? I once deleted a database instead of detaching it. Good thing I had backed up everything before I started working on it. Also I almost uninstalled a clients Exchange server because they connected me to the server and then told me I can uninstall anything I want since its not in production anymore. I sent a followup email explicitly asking about exchange since I was seeing packets still flying back and forth on it.
|
|
#
?
Sep 24, 2015 22:32
|
|
|
My test and production are exact clones, so that also means same db's. I've gone an edited the shell login scripts to say TEST at the end of the prompt. It's a nice, constant reminder of where I am.
|
|
#
?
Sep 24, 2015 22:35
|
|
|
My AWS file upload page is working for multi-part uploads with user logins with varied permissions. It'll be ready to go live after some chroming and stress testing. I did a thing, guys 
|
|
#
?
Sep 24, 2015 22:37
|
|
|
larchesdanrew posted:"My computer is really slow and Premiere crashes when I try to open it. " "Ok lets have a look at treesize to have gander at our files... why the gently caress has this guy got several tens of gigabytes worth of a my documents folder?" Filled to the brim with pirate movies and TV shows, I told him to get that off the company laptop and he would be in for a world of poo poo if anyone else knew.
|
|
#
?
Sep 24, 2015 22:42
|
|
|
Our dev and production databases are identical, but on completely separate hosts. Thankfully we haven't had any issues with anything in production, although one time I accidentally changed the email address of about 400 test accounts to the same one with a bad where clause Worst production issue was probably taking down every production web host at the same time instead of staggering them even/odd like they were supposed to, resulting in about 5-10 minutes of hard downtime. The nice thing about companies that don't poo poo on you for stuff like that is that I ended up being harder on myself than my boss ever would have been, and I rewrote a bunch of production deploy tools from scratch to make sure I don't screw stuff up like that again. Worst issue from my team, though, was a cron job that was sending out money via a payment provider. Not to go into super detail, but the provider was taking longer to confirm transactions than the cron's interval. Server 1 would find an unfilled invoice, associate and send payment for it, and the expectation was that the constantly-running handler on Server 2 would iterate through sent-but-not-confirmed payments, see that one was confirmed, find the associated invoice and mark everything as complete. Instead, Server 1 found an unfilled invoice, put the invoice number on a payment, fired off the payment... then on the next cron interval, found the same unfilled invoice, put the invoice number a (new) payment, fired off the payment... That one was pretty bad.
|
|
#
?
Sep 24, 2015 22:50
|
|
|
A client of ours has decided to reduce their reliance on Microsoft Office (for what I can only assume to be purely ideological reasons) by just not buying more licenses when they purchase PCs. That's literally as far as the plan goes.
|
|
#
?
Sep 24, 2015 23:02
|
|
|
ilkhan posted:I...D....E....? Then for a few years HP shipped PCs with SATA hard drives and IDE disc drives for some bizarre reason, we still have some of those floating around (not in my area fortunately).
|
|
#
?
Sep 25, 2015 00:09
|
|
|
So I'm having some issues with 6800 IA switches. The design requires us to run 15.2 in order to leverage the increased number of switches per stack (5) and the increased number of total switches supported per 6880 (42). I can provision and join each FEX stack while the 6880 is running 15.1, but the switches won't auto upgrade or join while the 6880 is running 15.2. Any ideas? e: Woah, wrong thread, but I guess I'll keep it here too. psydude fucked around with this message at 01:23 on Sep 25, 2015 |
|
#
?
Sep 25, 2015 00:37
|
|
|
Ozz81 posted:DING DING Would I be able to pick your brain about that script? I work for an MSP and we get tards clicking on "resumes" all the time. Most of it is remedied by rolling back the VSS snapshots but still... It would be awesome to clamp everything down in the event of an infection.
|
|
#
?
Sep 25, 2015 02:30
|
|
|
Knormal posted:Then for a few years HP shipped PCs with SATA hard drives and IDE disc drives for some bizarre reason, we still have some of those floating around (not in my area fortunately). My last few comps I built with CD/DVD drives were this way, just because there were usually only 2 SATA ports, plus an IDE, and a disc drive doesn't really gain anything from the better bus.
|
|
#
?
Sep 25, 2015 06:37
|
|
|
Thanks Ants posted:A client of ours has decided to reduce their reliance on Microsoft Office (for what I can only assume to be purely ideological reasons) by just not buying more licenses when they purchase PCs. That's literally as far as the plan goes. Did someone print out this plan on double-sided paper and someone only photocopied the front?
|
|
#
?
Sep 25, 2015 08:47
|
|
|
larchesdanrew posted:Oh well, at least I'm able to sit down for at least a few minutes and enjoy my coffee. I go to pick it up and the lid falls off and the cup goes sailing to the floor. This is why you drink coffee out of an actual coffee cup on your two 15-minute coffee breaks per day. Coffee isn't supposed to be mobile and the only type of drinking vessel that's supposed to have a lid is this:  tl;dr: You're doing coffee wrong.
|
|
#
?
Sep 25, 2015 10:39
|
|
|
|
|
#
?
Sep 25, 2015 11:33
|
|
|
You know who drank out of a lidded tankard? Hint: he was also big on living and dying for the fatherland.
|
|
#
?
Sep 25, 2015 11:43
|
|
|
Jerry Cotton posted:You know who drank out of a lidded tankard? Hint: he was also big on living and dying for the fatherland. Wilhelm the Second?
|
|
#
?
Sep 25, 2015 12:05
|
|
|
 Pictured: typical mug user.
|
|
#
?
Sep 25, 2015 12:33
|
|
|
Ghostlight posted:
There, fixed that for ya. 
|
|
#
?
Sep 25, 2015 13:23
|
|
|
Stupid things I've done? Set up automatic ticket creation via email without accounting for Out-Of-Office/Automatic replies.
|
|
#
?
Sep 25, 2015 13:25
|
|
|
JohnnyCanuck posted:Stupid things I've done? gently caress I have a user like this. Every time you email him, you get this huge written response about how you've reached the literal Adonis of anchors and he just can't wait to respond to your email and thanks for watching and blah blah blah, like people are writing to his loving fan club. He's the guy that submits all the "I can't see poo poo" tickets. At least three a day, and every time a ticket is opened, commented, and closed, it emails the submitter. So he submits a ticket, gets a confirmation email, auto responds which creates a comment which sends an email which auto responds which creates a comment which sends an email over and over and over until the email server just gives up. I turned that poo poo off real fast for him.
|
|
#
?
Sep 25, 2015 13:33
|
|
|
Reminds me of the infamous Bedlam Incident: http://blogs.technet.com/b/exchange/archive/2004/04/08/109626.aspx
|
|
#
?
Sep 25, 2015 14:03
|
|
|
|
| # ? May 17, 2024 19:33 |
|
|
I kinda want to use one for a coffee cup, just for the looks I'll get.
|
|
#
?
Sep 25, 2015 14:15
|
|