|
wolrah posted:Remember how in Windows 9x and I think maybe XP if you had enabled the network tray icon double-clicking on it would immediately get you to the equivalent of this dialog? I will never forget to ncpa.cpl
|
# ? Jun 5, 2020 17:02 |
|
|
# ? May 25, 2024 05:34 |
|
xzzy posted:Giving servers DHCP addresses has always made me nervous. While I realize that not everybody has a secondary site they can put the failover server in, but MS DHCP failover works really well. Except for the part where you have to set up and configure a power shell script to run in the background to copy over any reservation changes, etc.
|
# ? Jun 5, 2020 18:21 |
|
Pissing me off: My company being "cute" with their all hands memos.quote:... As you know, we committed to maintain employment for all employees until at least July 1st. We no longer are able to keep that commitment. I mean, it's great that none of the non-billable folks are getting laid off, but way to bury the lede.
|
# ? Jun 5, 2020 18:41 |
|
zharmad posted:Pissing me off: My company being "cute" with their all hands memos. this is psychotic
|
# ? Jun 5, 2020 18:48 |
|
i'm afraid we're going to have to let you go... to the employee appreciation party!! ha ha. why are you crying
|
# ? Jun 5, 2020 18:51 |
|
Pissing me off: Over-the-wall cube neighbor CONSTANTLY loving CLICKING TO REFRESH THE TICKETING SYSTEM
|
# ? Jun 5, 2020 19:24 |
|
SlowBloke posted:Everyone i've meet that has been vocal for teams to have multiple chats/channels in tabs or popup windows has been a trillian or pidgin user, i wonder what's the link 🤔 As I recall, Trillian did that because AOL instant messenger did that. So basically, in 2020, people still want loving AIM.
|
# ? Jun 5, 2020 19:27 |
|
zharmad posted:Pissing me off: My company being "cute" with their all hands memos. https://www.youtube.com/watch?v=0BUBd9dQvtY
|
# ? Jun 5, 2020 19:38 |
|
xzzy posted:Giving servers DHCP addresses has always made me nervous. In practice it's probably fine 99% of the time but I really don't want to hit that 1% where I get paged in the middle of a weekend because the DHCP server took a poo poo and some server went offline because it decided to be a good citizen and stop using an IP it no longer has a lease for (and researching the behaviors of all the various devices/operating systems out there is not easy). At work we don't use static reservations, we have to set it locally. Usually. One time I requested a static for a machine and Networking said the static range on that floor was full. But good news ! They'd punch a hole in the dynamic range and give that one machine a static assignment. Possibly the only static IP on campus being managed right (to my knowledge), and it's still being implemented wrong.
|
# ? Jun 5, 2020 19:41 |
|
If Arrested Development was all this type of stuff and not.. all the rest, I'd definitely have liked it more. zharmad posted:Pissing me off: My company being "cute" with their all hands memos. There's no way I'd let them go without a bollocking for that. It's just irresponsible.
|
# ? Jun 5, 2020 19:45 |
|
xzzy posted:Giving servers DHCP addresses has always made me nervous. In practice it's probably fine 99% of the time but I really don't want to hit that 1% where I get paged in the middle of a weekend because the DHCP server took a poo poo and some server went offline because it decided to be a good citizen and stop using an IP it no longer has a lease for (and researching the behaviors of all the various devices/operating systems out there is not easy). Ok, sit down and hear this. I worked at a small start up that had two different internet connections with two different LANs (with the same IP address scheme) and two wifi APs (one on each network). Both AP's had the same SSID and password. They also had a few wireless printers, it wasn't clear which printers were on which network, and which printer was even named what, (they were several of the same kind) so when you'd print it was a crapshoot which network you were on and which printer you got. It was a nightmare.
|
# ? Jun 5, 2020 21:03 |
|
Jaded Burnout posted:There's no way I'd let them go without a bollocking for that. It's just irresponsible. Don't get me wrong, its honestly a great company to work for. When covid started they gave all employees an automatic 80 additional hours of PTO for use for any reason, and made the commitment in early March to not lay anyone off. They also changed the training policy from 4 hours of training a month to unlimited training hours per pay period (2 periods/month) with the only requirement that the training be relevant to your job or in line with your career goals as discussed with your supervisor. Since they also stopped all travel through September, that freed up ~$30 million to shift to employee pay. They also added an additional $500-2,000/month to all family care FSA's from the firm to help ensure people didn't lose daycare slots even if they couldn't send their kids. I just have no idea what personal assistant or C-level would have approved the drat wording of that memo. It's so out of character for communications from our senior execs.
|
# ? Jun 5, 2020 21:13 |
|
Jerk McJerkface posted:Ok, sit down and hear this. Sounds like great redundancy!
|
# ? Jun 5, 2020 21:56 |
|
xzzy posted:Giving servers DHCP addresses has always made me nervous. In practice it's probably fine 99% of the time but I really don't want to hit that 1% where I get paged in the middle of a weekend because the DHCP server took a poo poo and some server went offline because it decided to be a good citizen and stop using an IP it no longer has a lease for (and researching the behaviors of all the various devices/operating systems out there is not easy). Beyond all that, realistically how often have you actually had DHCP break? I honestly can't think of a time I've had a DHCP server stop working unexpectedly. Sure, if I'm tinkering with configs I've broken 'em plenty but I figure that out pretty quickly when I test my change. And obviously if the hardware fails and the entire machine goes down, but I'm talking about just a silent failure of DHCP that would only be noticeable when leases start to run out or new clients fail to connect. I've never really felt like concerns about DHCP reliability were really all that legitimate. --- On the opposite end, a very nice advantage of using reservations rather than hard-coded statics is that it's self-documenting. It's impossible for it to end up out of date or for a typo to result in a server being listed at one address but having another. The fact that it works proves the documentation to be correct. Think about how many times you've encountered an IP address conflict caused by someone setting an address they weren't supposed to or not documenting what they've used. It's orders of magnitude more common than DHCP failures, that's for sure. --- My personal rule is that actual hardcoded static IPs should only be used for the DHCP server(s) and any devices that need to have an IP to allow the DHCP server(s) to do their job. DC and DNS in Windows environments, any routers that may be acting as DHCP relays for distant clients in larger networks, etc. Anything else just gets a reservation. Also anything that does truly need a hardcoded static still gets a reservation for documentation purposes. wolrah fucked around with this message at 00:23 on Jun 6, 2020 |
# ? Jun 6, 2020 00:18 |
|
wolrah posted:Beyond all that, realistically how often have you actually had DHCP break? I honestly can't think of a time I've had a DHCP server stop working unexpectedly. Sure, if I'm tinkering with configs I've broken 'em plenty but I figure that out pretty quickly when I test my change. And obviously if the hardware fails and the entire machine goes down, but I'm talking about just a silent failure of DHCP that would only be noticeable when leases start to run out or new clients fail to connect. Exactly once. Someone left an mmc console open on a remote session and it somehow consumed all the memory of the VM.
|
# ? Jun 6, 2020 01:05 |
|
Ah DHCP. Anyone ever take a basic network cert test that had one question where it's a home office and some satellite offices, and the question is about where DHCP and DNS should be located ? I worked for a company that got that wrong. It was the ad agency of course. My office was in San Francisco, and one weekend I'm at a buddy's place in Concord (about 35 miles as the crow flies) grilling and getting a good buzz on. Fortunately it was early when my cell rings. Everything is down. Shits on fire, and there's a new business pitch on Monday. Anybody who left their machine on Friday night can get to the file server but nothing else, even email. Nobody who came in with a laptop or started their machine up can connect to anything. That's a network thing, so I called the on-call. It turns out that's the weekend when they shut down the LA datacenter to cut over to the new electrical system. Literally everything is offline and powered down, and will be until Sunday night. No, we didn't get notified because we didn't use any services hosted in LA. Except DHCP, DNS, email, and FTP. Sucks to be us. I called the office back and said they were hosed until after 8pm on Sunday, and under no circumstances should anyone who could connect to the file server reboot. Monday our Groupwise guy took 10 minutes out of his busy schedule and spun up DHCP and DNS on our file server. Nobody involved got the slightest bit of heat for that, because "gently caress SF" was an actual policy.
|
# ? Jun 6, 2020 01:55 |
|
Correct answer is at each place right?
|
# ? Jun 6, 2020 01:57 |
|
Highly depends on your network architecture and redundancy. When I worked for a retailer we didn’t have anything local to the store except for a server for video surveillance and a one-way DFS replication share. But, each store was on an MPLS circuit with cellular failover.
|
# ? Jun 6, 2020 02:16 |
|
devmd01 posted:Highly depends on your network architecture and redundancy. When I worked for a retailer we didn’t have anything local to the store except for a server for video surveillance and a one-way DFS replication share. But, each store was on an MPLS circuit with cellular failover. Redundancy ? LOL. We were on a fractional T1 frame relay connection back to LA for everything, including Internet. Local management screamed for years to upgrade the Internet connection. Nothing happened until the HR manager from LA visited and emailed everyone an 8 MB .ppt. 8 meg file goes up the narrow pipe to the mail server. All cool and good. 70 POP3 clients try and download an 8 MB file more or less simultaneously. poo poo's broke and on fire. It took an hour to get enough people deleting the message in webmail that traffic died down to the point where getting the rest of the office to nuke the message only took 40 minutes. A couple of people actually got the download. So that was about two hours of the next best thing to a complete network outage. Mondays in advertising. Guess what she did on Tuesday ? Only took about 75 minutes to clean up, people knew the loving drill.
|
# ? Jun 6, 2020 03:34 |
|
Pro: All our devices use DHCP reservations, even if they are set to static there's a DHCP reservation configured. No dynamic DHCP except for defined ranges for guests or the few laptops that may visit a different location. The reservations are our single source of truth regarding network documentation, who owns which device, its purpose, etc. Con: DHCP is managed by an appliance and its API is barely there and I'm writing a loving web scraper so I can export the reservations to use elsewhere… I kind of want ISC DHCPD back…
|
# ? Jun 6, 2020 12:04 |
|
Heh, look at all you fancy lads with your DHCP. Here I am still administering thousands of devices that use a combination of bootp/tftp. My favorite so far was having to modify bootp itself so it would actually run on a server with an OS newer than 15 years old. I tried for a very long time to get them to talk to DHCPD properly, but gave up when re-automating everything would have been a bigger lift than just refactoring bootp. The day we turn these devices off, I'm throwing a party.
|
# ? Jun 6, 2020 17:28 |
|
Netbooting is so loving dumb. It should be so simple, but it's hilariously stupid. Dumb. Idiotic. Seriously. So dumb. Whether it's the boot menu crap Apple grafted onto DHCP, straight bootp, the barely documented Intel boot agent vendor extensions, the idiotic complications around UEFI (lmao PE headers) and BIOS… And then there's stuff like PXEBS (pxe bullshit). Let me quote from the ipxe docs: quote:PXE boot server discovery is a hideously ugly concept; one of many such concepts to be found in the PXE specification. If you don't already know what it is, then you are lucky enough to almost certainly not need to worry about it. gently caress PXE. (I actually have commits in ipxe… )
|
# ? Jun 6, 2020 23:08 |
|
Pxeboot has been pretty reliable for me over the years, other than some inconsistency in firmwares where some of them configure the interface with the first packet and others wait like thirty seconds to harvest offers before choosing one. Would rather deal with the quirks than shuffling a usb drive between servers. gently caress provisioning hundreds of nodes without full automation.
|
# ? Jun 7, 2020 03:47 |
|
dragonshardz posted:What would you use instead? My org uses #separation to mean both "died" or "found another job" but "termination" to mean exclusively "fired" unless used in a benefits context Euphemisms are poo poo
|
# ? Jun 7, 2020 07:08 |
|
mllaneza posted:At work we don't use static reservations, we have to set it locally. Usually. One time I requested a static for a machine and Networking said the static range on that floor was full. But good news ! They'd punch a hole in the dynamic range and give that one machine a static assignment. Possibly the only static IP on campus being managed right (to my knowledge), and it's still being implemented wrong. DO WHAT??
|
# ? Jun 7, 2020 07:09 |
|
Am I trippin or does the "new admin email" function in WordPress send the confirmation email to the newly entered email address? Seems like a bigger security concern than sending it to the existing email address. I mean... it's stupid enough... but then again WordPress is a pig of 20 developers spaghetti code so who knows what security issues I will find this week... Last week found a user elevation bug in a private plugin... a dozen regular plain jane users were site admins without them even knowing.
|
# ? Jun 8, 2020 16:51 |
wolrah posted:Remember how in Windows 9x and I think maybe XP if you had enabled the network tray icon double-clicking on it would immediately get you to the equivalent of this dialog? I was about to say "maybe it would be smarter to just learn the Netsh tool to look up and manage network stuff", but then... code:
|
|
# ? Jun 8, 2020 21:59 |
|
stepped in to lend a hand with tickets for an hour today, because half our team is out and we only had 1 guy on frontline. in the space of that hour, I found: 3 separate automation issues that, instead of spending the time to fix, the juniors instead had setup scheduled jira tickets to manually cleanup the automation failures. one of them i literally fixed in 5 minutes of looking at the job bc it was literally just pointing to the wrong group. i also had to seriously rolleyes at our more competent junior too. generally I like him but did everyone just forget their loving eyes at home today? a user was logging in with saml to our HR managed training system. they were getting the error "there are multiple accounts in our system associated with this email address". HR just saw a red box and said "put in an IT ticket" and our junior's solution to solving it was... unassigning and reassigning the okta user. When that didn't work, he tried swapping between individual and group based assignments. ...none of which has anything to do with the error she was getting! It was obviously an issue with the application, not Okta. individual vs group assignments change the user attributes sent in the SAML assertion but don't otherwise affect login flow at all. it was just clearly a case of people throwing poo poo at the wall to try and see what sticks instead of reading the loving error on their screen and going through the data flow. It ended up being fixed in two minutes when I asked HR to look at her account in the app and remove the duplicate account. i also feel like I talk about data flow a lot but I also don't see how you can understand anything in this industry without understanding how data is flowing. management must be hell it's just delegating to people who do things worse than you The Iron Rose fucked around with this message at 22:31 on Jun 8, 2020 |
# ? Jun 8, 2020 22:09 |
|
The Iron Rose posted:it was just clearly a case of people throwing poo poo at the wall to try and see what sticks instead of reading the loving error on their screen and going through the data flow. I feel this pain every day. The Iron Rose posted:i also feel like I talk about data flow a lot but I also don't see how you can understand anything in this industry without understanding how data is flowing. Also this.
|
# ? Jun 8, 2020 22:16 |
|
Ritualistic IT. "I do this dance and then the thing is usually fixed! Why? No idea, but it works!" Then as you get higher up, cargo culting. My god, so much cargo culting.
|
# ? Jun 8, 2020 23:03 |
|
quote:The Machine Spirit guards the knowledge of the ancients. Flesh is fallible, but ritual honours the Machine Spirit. To break with ritual is to break with faith.
|
# ? Jun 8, 2020 23:08 |
|
How we'll all look one day:
|
# ? Jun 8, 2020 23:30 |
|
We use departed, but I could also get behind "former". Meanwhile, I think my new minion has that thing where he sees sounds and hears colors, except as it relates to root causes. Every time I talk to him about an issue, it's "this smells like cache and cookies" or "hm, this doesn't taste like DNS." Not only is this disgusting, he's annoyingly frequently right with his hunches...
|
# ? Jun 8, 2020 23:50 |
|
Do the right thing. grab a whiteboard, Start tracking hunches for reliability, and start an office pool
|
# ? Jun 9, 2020 00:06 |
|
Sounds like he’s got ISP
|
# ? Jun 9, 2020 02:43 |
|
Twat Waffle posted:How we'll all look one day: Dafuq? What is the origin of this? A sci-fi RPG I’m guessing, but wow.
|
# ? Jun 9, 2020 05:07 |
|
A cogboy (Tech-Priest) of the Adeptus Mechanicus, in Warhammer 40,000. Where technology is fixed by following sacred rites and procedures.
|
# ? Jun 9, 2020 05:10 |
|
I've never felt more at home in a game than in Warhammer 40 000: Mechanicus. This, I decided, is my calling as a sysadmin. Seriously, the game is shockingly good.
|
# ? Jun 9, 2020 07:59 |
|
The Iron Rose posted:management must be hell it's just delegating to people who do things worse than you lol I might actually use this quote in future when I have to explain why I don’t want a promotion. too accurate Thing getting on my nerves today: new, indecisive lead lets the team debate and estimate really simple tickets for ages. I long for a draconian planning system where everyone only gets to say one sentence per ticket.
|
# ? Jun 9, 2020 08:23 |
|
|
# ? May 25, 2024 05:34 |
|
Antigravitas posted:I've never felt more at home in a game than in Warhammer 40 000: Mechanicus. This, I decided, is my calling as a sysadmin. Looks pretty good yeah.
|
# ? Jun 9, 2020 14:49 |