|
rafikki posted:What does that say about the quality of your emails?
|
# ? May 9, 2023 20:50 |
|
|
# ? May 25, 2024 20:23 |
|
If I was Sirotan would just send that lovely avatar as my emails everywhere it makes me happy every time I see it.
|
# ? May 9, 2023 22:13 |
|
rafikki posted:What does that say about the quality of your emails? Lmfao
|
# ? May 9, 2023 22:14 |
|
god drat what a burn
|
# ? May 10, 2023 00:20 |
|
Internet Explorer posted:god drat what a burn Wish there was a good way to turn it into a thread title.
|
# ? May 10, 2023 01:46 |
|
pretend i posted the bernie sanders meme saying "i am once again asking you to update the thread title" without having to figure out this imgur apocalypse workaround
|
# ? May 10, 2023 01:50 |
|
Rick posted:If I was Sirotan would just send that lovely avatar as my emails everywhere it makes me happy every time I see it. rafikki got me good, though the thing I emailed myself might be of interest to other folks here: the link to a Facebook class action so I wouldn't forget to file a claim to get $0.63 five years from now https://facebookuserprivacysettlement.com/#submit-claim
|
# ? May 10, 2023 02:34 |
|
Yes, call me six times every 2 minutes an hour before I'm on call. Do not leave a message, this will definitely endear me to giving a poo poo.
|
# ? May 10, 2023 10:00 |
|
Cleaning up more fun messes in elasticsearch! Today our problem child is network flow logs. Literally just “SOURCE IP X.X.X.X DESTINATION IP Y.Y.Y.Y” from AWS VPC flow logs. Our security team and one of our infrastructure guys threw this together by logging everything to s3, having our logging parsing system read from there and post to elasticsearch. Only problem is they did this without any monitoring, performance considerations, or data handling. Just casually dumped it into elastic. This consisted of about a billion (literally, not figuratively) separate documents per day, making up 98% of the total volume of logging ingest we do on a daily basis. 125TB of required storage capacity across the 90 day data lifetime when our current cluster max is 45TB. No custom index template so we had 100gb shards. When you take into account memory requirements, handling this data would require a cluster twenty six times larger than we have currently. we *already* spend roughly 10k/month on infrastructure costs for this system, and that’s excluding licensing costs (~250k/yr) which are also predicated on size. anyways needless to say the project is now dead, my cluster is much happier and no longer sending tcp 0 windows due to backpressure 5-10x every minute, and we all learned a valuable lesson about checking with your SRE before you start handling big data. The Iron Rose fucked around with this message at 18:44 on May 10, 2023 |
# ? May 10, 2023 18:30 |
|
Sec Ops promising to automate my patch deployment and then me ending up with a bunch of Security Update - Security Update Security Update - Security Update Security Update - Security Update Security Update - Security Update Security Update - Security Update Security Update - Security Update on my SCCM deployments without a goddamn hint of what and where things are being sent without drilling in, very nice, more AI everywhere please
|
# ? May 10, 2023 19:16 |
love being goaded into updating to visual studio 2022 when about once a month the solution to "why doesnt this extension work" is "vs 2022 isn't supported, downgrade to vs 2019"
|
|
# ? May 10, 2023 19:49 |
|
The Iron Rose posted:anyways needless to say the project is now dead, my cluster is much happier and no longer sending tcp 0 windows due to backpressure 5-10x every minute, and we all learned a valuable lesson about checking with your SRE before you start handling big data. The bolded part is unreasonably optimistic.
|
# ? May 10, 2023 20:11 |
|
ElGroucho posted:Sec Ops promising to automate my patch deployment and then me ending up with a bunch of
|
# ? May 10, 2023 20:13 |
|
Wizard of the Deep posted:The bolded part is unreasonably optimistic. Yeah, no joke there. The problem described is the same problem as poo poo from 20 years ago with different product names. An issue I've been dealing with for the past week, with tons other engineers spending countless engineer hours, ends up being "some applications aren't optimized to communicate to a database across a WAN and introducing that latency is going to result in poor performance." When I heard "we are migrating a database and performance is bad" I knew exactly what the issue was, I'm sure everyone else did too, and yet here we are.
|
# ? May 10, 2023 21:17 |
|
The Iron Rose posted:words This makes me so happy I dont have to admin Elasticsearch anymore.
|
# ? May 10, 2023 21:30 |
|
Polio Vax Scene posted:love being goaded into updating to visual studio 2022 when about once a month the solution to "why doesnt this extension work" is "vs 2022 isn't supported, downgrade to vs 2019" You should be updating to a Jetbrains IDE.
|
# ? May 10, 2023 21:32 |
|
FlapYoJacks posted:You should be updating to a Jetbrains IDE. Just use Vim and compile in the terminal.
|
# ? May 10, 2023 21:34 |
|
Hughmoris posted:Just use Vim and compile in the terminal.
|
# ? May 10, 2023 21:39 |
|
BaseballPCHiker posted:This makes me so happy I dont have to admin Elasticsearch anymore. I just resolved a new problem, which at least I was actually responsible for, that killed about 75% of new logs for our primary service because we had an explosive dynamic field type explosion. Our parser had an api_params field with several thousand different possible fields. It’s been dropped now, folks can just use the inverted index and partial string matches on the full json blob in “message” if there was anything there they actually cared about. that’s at least actually indexed as a text field. Still need to update a bunch of field mappings so we can do minor things like, idk, partial matches on kubernetes pod names, and then reindex everything, and then there’s the type conflicts in the data view and aaaaaghhhhhhhh gently caress elasticsearch is what I’m saying.
|
# ? May 10, 2023 22:16 |
I spent about 30 min today making GBS threads on elasticsearch in a client meeting lmao fuckkkkk running it yourself
|
|
# ? May 11, 2023 01:24 |
|
KillHour posted:The fraud person said they had the last 4 of my credit card and my CVV code, so they're sending me a new card. Wonder how they got it. Its likely you were compromised from somewhere during the april Genesis Market shutdown. The day after the raid went down i had my credit card frozen after someone tried to spend 20$ on some lovely stealthwriter AI thing. (all it takes is one lovely merchant saving things in plaintext for example, a camera/skimmer at a store, or other incompetence) source This was a month ago in april, *but* it put a time limit on bad actors who bought the stolen accounts to use it before everyone went 'oh poo poo better change everything'.
|
# ? May 11, 2023 02:19 |
|
not only do I have to ingest all that flow data (at least we only need to keep 14 days now), but I have to spin up a new EKS cluster and migrate a 15TB elasticsearch cluster because our we ran out of IP space in the existing subnet. EKS, the worst managed kubernetes service, doesn’t let you expand or add new subnets to existing clusters*. *it does but you don’t get cluster ENIs The Iron Rose fucked around with this message at 16:54 on May 11, 2023 |
# ? May 11, 2023 16:34 |
|
we just pay for data dog
|
# ? May 11, 2023 16:36 |
|
I’m just going to take downtime, size up the existing node pool, and then move our company out of AWS into GCP and Azure to avoid the inevitable scaling cap.
|
# ? May 11, 2023 16:45 |
|
conversation I just had: dev: my app won't deploy to prod me: what is the error message? dev: connection timeout me: what agent pool are you using? dev: non-prod
|
# ? May 11, 2023 18:08 |
|
Why does SharePoint suck so much? We have a client we're looking to migrate from Citrix ShareFile; one of the main requirements is that they're able to retain relative links to PDFs in Excel cells, which works just fine at the moment since the ShareFile sync app simply maps/acts like a local S: drive. While testing with an example .xlsx file located in a synced SharePoint library, the hyperlinks open in a web browser rather than the locally-installed Acrobat desktop application, because it's basically just a web URL now. There's no way around this as far as I can tell, with the only exceptions of course being "supported Office file types" being able to open in their native Word/Excel/PowerPoint desktop versions. gently caress off, Microsoft!
|
# ? May 11, 2023 22:43 |
|
I deploy some new Meraki APs to a site, they are the first to get them outside of our lab. I ask people onsite to give me feedback, test it a bunch and just let me know how it works. I hear nothing from anyone there. Then today, I get a email from the IT director with a note from someone there complaining about the wireless. Instead of telling me, they instead just took their bitching right to the top. Nevermind we were forced to send these out before testing was done, and we even had test environment that wasn't a lab.
|
# ? May 12, 2023 20:17 |
|
MiniFoo posted:Why does SharePoint suck so much? We have a client we're looking to migrate from Citrix ShareFile; one of the main requirements is that they're able to retain relative links to PDFs in Excel cells, which works just fine at the moment since the ShareFile sync app simply maps/acts like a local S: drive. While testing with an example .xlsx file located in a synced SharePoint library, the hyperlinks open in a web browser rather than the locally-installed Acrobat desktop application, because it's basically just a web URL now. There's no way around this as far as I can tell, with the only exceptions of course being "supported Office file types" being able to open in their native Word/Excel/PowerPoint desktop versions. gently caress off, Microsoft! If you contact Microsoft, they’ll explain that SharePoint is integrated primarily with Office Online. As a former SharePoint Online admin, my advice is to go with the flow because fighting Microsoft’s model is a pain in the rear end.
|
# ? May 12, 2023 21:03 |
|
FlapYoJacks posted:You should be updating to a Jetbrains IDE. This brings back memories of being one of the Teamcity admins for our entire development pipeline.
|
# ? May 12, 2023 21:06 |
|
I'm working with a VR headset and I'm checking out a feature that allows the headset to quickly establish tracking and position in a space with the assistance of an ArUco marker. I think it's working? It seems to do what I want it to do, but the system gives absolutely no feedback. No feedback when setting it up whether it's been done correctly. No feedback on whether it is scanning for it. No feedback whether it has succeeded or failed. It works consistently and I'm seeing the expected behaviour when for example the ArUco marker isn't visible but when it comes into view, then the tracking is established. But there's nothing on the UI, nothing in the logs that confirms that the system is doing what I think it's doing. Fragrag fucked around with this message at 10:29 on May 23, 2023 |
# ? May 23, 2023 10:27 |
|
Fragrag posted:It works consistently and I'm seeing the expected behaviour when for example the ArUco marker isn't visible but when it comes into view, then the tracking is established. But there's nothing on the UI, nothing in the logs that confirms that the system is doing what I think it's doing. Plausible deniability
|
# ? May 23, 2023 15:14 |
|
Well I confirmed it was working by looking at an identical marker in a different location and it launched me into space. Real fun
|
# ? May 24, 2023 08:22 |
|
I found out today that our service desk/call center people got a call about a user saying they had a beeping in their network closest, so they had the person unplug the UPS from the wall, and then plug the switch into the wall. Then hung up and transferred the ticket. At no point in the ticket was it mentioned that they did this. Also, they aren't supposed to do this. Not only that, the UPS that was beeping wasn't even on the network equipment.
|
# ? May 31, 2023 03:10 |
|
Almost as good as the time the cleaner unplugged the UPS to plug in his floor polisher. Management never locked the server room door, and learned absolutely nothing from it.
|
# ? May 31, 2023 06:49 |
|
ITSec taking deployment of AV and definition updates away from Systems. Because how hard can SCCM really be? I haven't had to actively work with SCCM in years but I still nurture a healthy fear and loathing of the platform. There are so many ways to break that stupid pig into removing custom policies and rerolling default settings with no exclusions to clients, which fucks with Developer workflows every time it happens. And since ITSec is incapable of following change control procedures, the whole of IT looks like complete dicks every time it happens. Also pissing me off: I haven't been disillusioned enough with my job to regularly look at or post in this thread in a long time. I am not happy to be here again. afflictionwisp fucked around with this message at 14:11 on May 31, 2023 |
# ? May 31, 2023 13:55 |
|
afflictionwisp posted:ITSec taking deployment of AV and definition updates away from Systems. Because how hard can SCCM really be? I haven't had to actively work with SCCM in years but I still nurture a healthy fear and loathing of the platform. There are so many ways to break that stupid pig into removing custom policies and rerolling default settings with no exclusions to clients, which fucks with Developer workflows every time it happens. And since ITSec is incapable of following change control procedures, the whole of IT looks like complete dicks every time it happens. Nobody sane of mind pushes AV updates from SCCM, you can set it up to have feature/quality with SCCM and AV using windows update and avoid nonsense.
|
# ? May 31, 2023 14:17 |
|
afflictionwisp posted:And since ITSec is incapable of following change control procedures, What do they think their job is?
|
# ? May 31, 2023 15:19 |
|
Bargearse posted:Almost as good as the time the cleaner unplugged the UPS to plug in his floor polisher. Management never locked the server room door, and learned absolutely nothing from it. I feel like that exact thing happens at every workplace. We supposedly had a switch stack that would go down most weekday evenings, turns out vacuuming was the crux of it. They did get told to not do that once it was unearthed. This was before my time, but I’ve heard it enough that it’s not just folklore.
|
# ? May 31, 2023 15:21 |
|
SlowBloke posted:Nobody sane of mind pushes ... from SCCM I've been primarily storage services / vm hosting for near a decade now, so I just kinda watch from the sidelines. Seen so many people come and go on that platform. The lifespan of an SCCM admin seems to be about two years max from enthusiasm to complete burnout. I've tried to warn them. One of my fondest memories is a colleague coming to me and saying, "I understand now. You were right. I thought you just hated your job, but this thing is loving awful!"
|
# ? May 31, 2023 15:59 |
|
|
# ? May 25, 2024 20:23 |
Arquinsiel posted:
The job of IT Security? Of course their job is to obstruct business, and skipping change controls is a great way to aid that goal. (Part of this post may be in jest.)
|
|
# ? May 31, 2023 16:22 |