--- Log opened Thu Jan 30 00:00:49 2020 06:22 <@Evilpig> after months of firewall upgrades/failovers causing us heartburn on nfs sessions getting hung up. it looks like the issue with junos is the firewall pairs are out of sync on their session states. checking the id for my current process on the active device and the same session id on the passive device are completely different pairs of devices. :-/ 06:22 <@Evilpig> and they've been trying to tell us that this was working as intended this entire time 07:11 -!- dc0de [~jim@198.46.153.211] has joined #se2600 07:11 -!- mode/#se2600 [+o dc0de] by ChanServ 07:12 -!- dc0de [~jim@198.46.153.211] has quit [Remote host closed the connection] 07:16 -!- dc0de [~jim@198.46.153.211] has joined #se2600 07:16 -!- mode/#se2600 [+o dc0de] by ChanServ 07:32 < aestetix> I wonder how many broken tech systems are due to a system not being used in the way it was intended 07:33 < aestetix> such as a temporary system for a quick fix becoming over time a critical piece of infrastructure 10:00 <@Corydon76> aestetix: when's the last time you browsed http://thedailywtf.com ? 10:00 < PigBot> The Daily WTF: Curious Perversions in Information Technology (at thedailywtf.com) http://tinyurl.com/4geup 10:03 < aestetix> oh man it's been years 10:03 < aestetix> there's a system I set up ages ago as a makeshit solution to a short-term workaround 10:04 < aestetix> and now all this other stuff depends on it, and I have to figure out a way to upgrade it without breaking all the stuff 10:04 < aestetix> and of course the original purpose is no longer needed 10:08 <@Corydon76> aestetix: that's easy, as long as you have all the dependencies replicated in development and test environments. You have that, right? 10:09 < aestetix> heh. hah. yeah. of course. 10:09 <@Corydon76> You're not testing in production, right? 10:10 * aestetix sighs 10:13 < aestetix> See, this is what happens when "high priority" stuff keeps coming up and disrupting the workflow 10:15 <@Corydon76> That's kind of why I came in and implemented Nagios and wrote tests that are early warnings for problems. 10:16 <@Corydon76> For instance, I have a test that checks the last order entered into a system, which assumes that we get orders a regular basis. It occasionally alarms over holidays, which isn't a big deal, but it saves my ass when somebody realizes we haven't seen an order from a particular site for some period of time. 10:17 <@Corydon76> When I came on, we regularly had emergencies, where nobody really had a clue whether something was a recent emergency or if something had been broken for awhile, and just nobody noticed. 10:18 <@Corydon76> "Site is down!" And then I discover it's been down for 3 days, and nobody noticed until now. 10:20 <@Corydon76> My current concept of fighting fires is along the lines of seeing all of the systems alert me that I've got X updates to perform, that nobody else thinks is an emergency until 3 days later, when the "normal" update turns out to be a quietly pushed out security update. 10:21 < aestetix> yep 10:21 < aestetix> my favorite updates in linux are the ones that demand you reboot to finish them 10:21 < aestetix> especially if they are not kernel related 10:22 <@Corydon76> Weird. I never see such. The only ones I see are kernel updates. Still a PITA, because if it's customer-facing, I pretty much have to do them at 2am. 10:23 <@Corydon76> Oh, wait, I do see those on the desktop, but desktop Linux machines aren't production equipment. 10:25 <@Corydon76> Shouldn't be, anyway. Fuck you, Red Hat, for letting clients put X environments on production servers. 10:27 < aestetix> why the hell would you have X on a server? 10:28 <@Corydon76> Because people are fucking idiots. 10:33 <@Corydon76> And also because if you're Red Hat certified, it assumes that you have the X environment and desktop installed. Their certification program includes next to nothing about how to do things like managing users on the command line. 10:34 <@Mirage> Waay back when K9 was still at VU he opened a ticket with NEC for the PBX reporting software. The NEC support tech, being the "I can't think for myself and can only recite what is in the document in front of me" type, informed him that on the RHEL server he would have to install X in order for them to proceed further with troubleshooting. After he installed X (as an emergency change) and got back in contact with them. His next instruction from NEC was " 10:36 <@Mirage> K9 just about lost his shit at that point...but not directly to support. 10:37 <@Corydon76> Mirage: next instruction was to open a terminal? 10:37 <@Corydon76> Screed was cut off 11:03 <@Mirage> yes 11:05 <@Mirage> He'd been SSH'd into the host the whole time because it was across campus, so he was just a little pissed and dismayed at them having had him install X just to open a terminal window when he could have done the same thing via the SSH connection he already had. 11:06 <@Mirage> Needless to say when that host was migrated from RHEL 4 -> 5, X was not installed again. 13:20 -!- rpifan [~rpifan@p200300D2673C27C5A60B7DF1DFFD1FE3.dip0.t-ipconnect.de] has joined #se2600 14:05 <@Mirage> damn girl scouts and their thin mint cookies 14:08 <@opticron> the ones that got me this time were smores 14:12 < xray> As in you eat one and want s'mores? 14:30 <@opticron> that's about the size of it 14:30 < rpifan> those cookies sucks 16:23 -!- rpifan [~rpifan@p200300D2673C27C5A60B7DF1DFFD1FE3.dip0.t-ipconnect.de] has left #se2600 [] 18:42 < K`Tetch_> all those girl scout ones do 21:03 <@Dagmar> Wow 21:03 <@Dagmar> This dude is proud he scored Viewsonic G90F monitors from his university 21:03 <@Dagmar> They just plain tricked him into getting them off their inventory 21:04 <@Dagmar> They're 19" (18" viewable lol) that max out at 1600x1200 and weigh _55lbs_ each 21:05 <@Dagmar> Roden: Could be worse 21:30 <@oddball> God, I remember those. --- Log closed Fri Jan 31 00:00:51 2020