|
MC Fruit Stripe posted:Nothing like sending an email first thing in the morning to let The Business know that execution of a massive project, scheduled for tonight, needs to be cancelled, and that it's My Fault. Ah the joy of "Would you rather hear this Thursday when half the company can't work, or today when I've prevented this problem?" My wife's office just sent an email about how their dropbox migration was postponed because a disk in their RAID5 failed and the IT person didn't want to stress the server out more before a replacement could be swapped in, fine. It then continues where he checks the status of their backups only to find the external usb disk failed weeks ago. Whoops. Thankfully they handled it as correctly as they could for missing that they had no functional backups for weeks, calling someone local to immediately go plug in a new disk and run a manual backup. Doesn't explain why it wasn't noticed, or why they don't rotate the disk offsite, etc.
|
#
?
Aug 2, 2017 15:15
|
|
|
|
| # ? May 25, 2024 21:42 |
|
|
Vargatron posted:Also, MC Fruit Stripe, godspeed. Better to pull the plug now than have the business brought to it's knees trying to push a project out the door too soon. I bet they'll make him deploy anyways, "accept" the risk of server not servering well enough and when everything comes to a crushing halt they will blame it on MC Fruit Stripe anyways. Most likely asking stupid questions like "how could we jot have foreseen this scenario!!?!?". I hope I'm wrong, good luck today comrade.
|
|
#
?
Aug 2, 2017 16:24
|
|
|
Sefal posted:My condolences Gnarly. One of the reasons I took this job is because I needed some time to decompress and slack off a bit. I appreciate the stability and easy workload every day though. I'm in the middle of building a house / selling a house / living out of boxes in some weird netherspace of home ownership. It'll be nice not to have to worry about poo poo when the wife and I are trying to have kids too. Compared to my last job, this poo poo is basically retirement... plus a 'part time, paint section, Lowe's job to kill time and keep me sane'.
|
|
#
?
Aug 2, 2017 16:46
|
|
|
GnarlyCharlie4u posted:Yes, no, and not here but in literally every other county it is for some reason. No. Keeping skills up is legit but interviewing is tiresome.
|
|
#
?
Aug 2, 2017 17:28
|
|
|
Today I'm getting third-party software patching working on our computer labs which use Deep Freeze. Using PDQ Deploy, and this guide. You can integrate "thaw" commands using a command-line utility that Faronix provides, run the patches/installations, and then refreeze them and reboot. Very easy and very cool. 
|
|
#
?
Aug 2, 2017 17:51
|
|
|
Been a long time since I used Deep Freeze, but we used to use it for our high school. It was a huge school and we had pretty much 0 tech resources, so they leveraged us kids a lot. A good portion of my time in school was spent in a randomly named class that was really just doing network admin stuff. We had it set up so we could thaw a machine, push an image with the enterprise version of Symantec Ghost, then have it come back up frozen. I agree, super awesome stuff and is really what got me into the concept of non-persistent machines. Now a days I design/deploy/administer non-persistent VDI infrastructure using XenDesktop. I was super fortunate to have that opportunity back in the day!
|
|
#
?
Aug 2, 2017 17:56
|
|
|
Internet Explorer posted:Been a long time since I used Deep Freeze, but we used to use it for our high school. It was a huge school and we had pretty much 0 tech resources, so they leveraged us kids a lot. A good portion of my time in school was spent in a randomly named class that was really just doing network admin stuff. We had it set up so we could thaw a machine, push an image with the enterprise version of Symantec Ghost, then have it come back up frozen. I agree, super awesome stuff and is really what got me into the concept of non-persistent machines. Now a days I design/deploy/administer non-persistent VDI infrastructure using XenDesktop. I was super fortunate to have that opportunity back in the day! I'm not directly involved with the labs but I bugged the guy who is since I do run our PDQ Deploy and I've been trying to get our Nexpose scores down. The labs are the low-hanging fruit and the only complication to patching them was their use of Deep Freeze, but this was surprisingly easy. I was hoping they'd convert to VDI but we just did a big expensive hardware refresh for them so it'll be a while, and probably not before I leave.
|
|
#
?
Aug 2, 2017 18:02
|
|
|
Today I learned something new. This is a thing: https://fogproject.org/
|
|
#
?
Aug 2, 2017 19:27
|
|
|
So I need to look into setting up a local server for the new site to temporarily handle things like AD/DHCP/DNS etc until everything gets moved over (Our ISP is looking and options to link both sites but there's not much worthwhile for the sake of a month), luckily I've got an older server computer I can use as a Hyper-V host so I get a quote for some server 2012r2 licensing and send it to the boss saying we need this. "Can we pay this on subscription or monthly?" Well this is Microsoft so maybe one day, but lol no 
|
|
#
?
Aug 2, 2017 20:44
|
|
|
Super Slash posted:So I need to look into setting up a local server for the new site to temporarily handle things like AD/DHCP/DNS etc until everything gets moved over (Our ISP is looking and options to link both sites but there's not much worthwhile for the sake of a month), luckily I've got an older server computer I can use as a Hyper-V host so I get a quote for some server 2012r2 licensing and send it to the boss saying we need this. Spin up an azure vm, connect your site via vpn, then shut it down at the end of the month. The azure compute price includes all licensing, including CAL's.
|
|
#
?
Aug 2, 2017 21:09
|
|
|
GnarlyCharlie4u posted:Today I learned something new. "COMPUTER CLONING AND MANAGEMENT" - okay, that's a start Activity is just code releases Features includes 'features' like being open source, having a forum, and OS support. There's a hint in scalability, "If you need to image just 5 computers or even 50,000 FOG can handle it" Okay so it's like a deployment tool I guess? Just, ugh, people, tell me what the hell your product is before you tell me you have a Wiki and a Github.
|
|
#
?
Aug 2, 2017 21:25
|
|
|
MC Fruit Stripe posted:God I hate the modern internet. I scrolled that entire page and I don't know what it does
|
|
#
?
Aug 2, 2017 21:26
|
|
|
anthonypants posted:It looks like Norton/Symantec Ghost, but free and open source. It also has an agent that allows limited endpoint management through a web console. It might be a useful imaging platform for a small/medium business that doesn't want to run Windows Servers, but if you have a full Windows infrastructure, you're better off using MDT.
|
|
#
?
Aug 2, 2017 21:30
|
|
|
Seems like a product that nobody needs. Baffling.
|
|
#
?
Aug 2, 2017 21:32
|
|
|
FOG has been around for at least a decade at this point. Back when your choice was Symantec Ghost or FOG, it wasn't a bad choice. I have no idea where it is now.
|
|
#
?
Aug 2, 2017 21:36
|
|
|
Sickening posted:Seems like a product that nobody needs. Baffling.
|
|
#
?
Aug 2, 2017 21:43
|
|
|
The Fool posted:Spin up an azure vm, connect your site via vpn, then shut it down at the end of the month. The azure compute price includes all licensing, including CAL's. I was hoping someone would say this! I was sort of aware Azure could do something like that but mostly heard mixed things about how it handles licensing, I have zero experience with it but I'll get my research and testing hat on.
|
|
#
?
Aug 2, 2017 21:45
|
|
|
Super Slash posted:I was hoping someone would say this! I was sort of aware Azure could do something like that but mostly heard mixed things about how it handles licensing, I have zero experience with it but I'll get my research and testing hat on. https://azure.microsoft.com/en-us/pricing/licensing-faq/ quote:The license to run Windows Server in the Azure environment is by default included in the per-minute cost of your Windows Virtual Machine. And quote:No. Windows Server CALs are not required for accessing Windows Server running in the Azure environment because the access rights are included in the per-minute charge for the Virtual Machines Things get more complicated when you have a hybrid environment, but for your use case, it's pretty simple. Depending on the size of your site, you could probably get away with running everything on a single D1 instance, which would run you about $100/month
|
|
#
?
Aug 2, 2017 21:55
|
|
|
MC Fruit Stripe posted:God I hate the modern internet. I scrolled that entire page and I don't know what it does From the makers of Zombocom.
|
|
#
?
Aug 2, 2017 22:10
|
|
|
The Fool posted:Spin up an azure vm, connect your site via vpn, then shut it down at the end of the month. The azure compute price includes all licensing, including CAL's. Roargasm fucked around with this message at 22:31 on Aug 2, 2017 |
|
#
?
Aug 2, 2017 22:22
|
|
|
GnarlyCharlie4u posted:Today I learned something new. THe niche for that is broke as gently caress libraries running computer labs running deep freeze. I used that for years and it is indeed awesome for it's niche.
|
|
#
?
Aug 2, 2017 22:35
|
|
|
Roargasm posted:We have a lot of problems running VPN nodes on the Azure network stack. The whole stack seems to objectively suck compared to AWS at high load. Speaking for linux stuff only, never tried RRAS it's also garbage.
|
|
#
?
Aug 3, 2017 00:52
|
|
|
I'm running a Very Significant Amount of Ubuntu on Azure with huge ingress and egress bandwidth usage, and I've never noticed a significant problem, whatever that's worth. What kinds of issues have you seen with VPN in particular (and what kinds of VPNs)?
|
|
#
?
Aug 3, 2017 02:08
|
|
|
It's been openswan and openVPN back to traditional datacenters, maybe 5 tunnels open. The whole thing falls over at like 400mbit, usually the nodes are unresponsive and have to be rebooted from console
|
|
#
?
Aug 3, 2017 02:32
|
|
|
Roargasm posted:It's been openswan and openVPN back to traditional datacenters, maybe 5 tunnels open. The whole thing falls over at like 400mbit, usually the nodes are unresponsive and have to be rebooted from console How does that compare to the routebased and policybased vpn options azure provides?
|
|
#
?
Aug 3, 2017 02:51
|
|
|
Naramyth posted:THe niche for that is broke as gently caress libraries running computer labs running deep freeze. I used that for years and it is indeed awesome for it's niche. Our library runs CleanSlate. Libraries are usually broke as gently caress considering the poo poo software I see vendors peddling. FOG seems like it'd be a great thing for LAN cafes and poo poo like that.
|
|
#
?
Aug 3, 2017 15:35
|
|
|
GnarlyCharlie4u posted:Our library runs CleanSlate. Libraries are usually broke as gently caress considering the poo poo software I see vendors peddling. I actually hadn't thought of that use case. That does seem like it would be good.
|
|
#
?
Aug 3, 2017 16:38
|
|
|
H110Hawk posted:Ah the joy of "Would you rather hear this Thursday when half the company can't work, or today when I've prevented this problem?" This syndrome irritates the hell out of me. When I was at Verizon, we had a released cancelled, and it irritated people so much than we spent the first ten minutes of our team meeting discussing it. That sounds like no big deal except that we were a UNIX sysadmin team and had no connection to the release - it had just been the subject of so much discussion that our director wanted to discuss it with us. About nine minutes in, I lost patience and said, "Why is this such a problem? We have testing, the testing found a problem, the release was cancelled. That's working as intended. If we did all that testing and never, ever found a problem, that would tell us that we had a poorly planned test cycle and were wasting time and resources on tests that never failed." The conversation petered out quickly after that. I would have thought it was a pretty obvious point.
|
|
#
?
Aug 3, 2017 18:50
|
|
|
Woof, nevermind thinking about linking these two sites this circus just got even better. A bunch of us went and did another site visit today plus I brought in an engineer to do a survey for the infrastructure, right off the bat we aren't allowed on site because the foreman hasn't had clearance plus we don't have PPE. So the boss works on clearance and somebody ropes in another company to give us just enough bunch of boots/vi-vis jackets/hardhats to get inside, while most of the crew bugger off and take photos my guy and a health & safety advisor get to work. The short short version of this is that it's a loving massive project for something that has to be completed by October and pretty much impossible, this isn't just an IT fit out into a new office this is literally constructing an office from bare bones. The ceiling is lovely for access to lay cable and despite everyone saying we need power poles the boss wont have it, server room space regs means the planned area will need to be split into two rooms to house a rack each, the electrical circuit is a hell of a job since it needs completely redoing, and there are reams of newer safety regs to comply with because of recent events (Today I learned Ethernet cable needs to be fire rated). So the engineer laid it all out and couldn't even soft ball the massive price tag behind this, so we'll need to come back next week for a better look at things with more professionals. When everyone left we both had a honest and humorous chat about how absurd this project will turn out and not doubt probably even cost six figures, I'm not sure if it was advice or bit of light heartedness but he pretty much said...  Personally I don't really care as it's waaaaay beyond my experience level, what's the SOP on new jobs? wait until you've got the new contract in hand before pulling the eject lever?
|
|
#
?
Aug 3, 2017 19:57
|
|
|
It's possible that your firm can go "look we're not going to touch this because the conditions are untenable".
|
|
#
?
Aug 3, 2017 19:59
|
|
|
How have you managed to get dragged into this?
|
|
#
?
Aug 3, 2017 20:33
|
|
|
Because I'm the sole "IT guy" in the company who manages the lot. At first I was asked to do the network and server build out which didn't sound too bad as I've got great relationships with provider companies to lean on (and I made it very clear they would be doing the bulk of the work), then I was asked to handle the electrical install as well which I questioned that's got nothing to do with IT. Fortunately my tech supplier can handle all of this as they have specialists who do this kind of thing, however this is my third chance to leave the company with a firm job offer on the table and I'm drat well taking it (turned downs help desk type job a couple years ago, and second was NOC support earlier this year).
|
|
#
?
Aug 3, 2017 20:50
|
|
|
Man if it gets you out of a hosed up situation I'd take the job.
|
|
#
?
Aug 3, 2017 20:57
|
|
|
mewse posted:The company we're taking over has on-prem active directory synced with office 365 via azure AD. mewse posted:You're probably right and I doubt this org was configured correctly. I'm just offering the rare bit of insight I've gleaned from my day job about why you were asked that question. Follow-up MS articles: This one describes the solution: https://support.microsoft.com/en-us/help/2492140/you-can-t-assign-a-federated-domain-to-a-user-in-the-office-365-admin This one describes how/why the software chooses the primary email address for users: https://support.microsoft.com/en-us/help/2392130/troubleshoot-user-name-issues-that-occur-for-federated-users-when-they After looking into it more, if you fill in the "email address" field for a new user in active directory before it syncs via azure AD, the correct primary email address will be chosen in office 365 and there's no need to screw around with the ProxyAddresses property.. but you do have to do that if the user exists and is misconfigured.
|
|
#
?
Aug 3, 2017 22:39
|
|
|
Just fixed an issue which has plagued us since yesterday morning. Took a lot of time and got almost no sleep, but I think most of you guys understand how good it feels to fix something like that. King of the world baby.
|
|
#
?
Aug 3, 2017 22:42
|
|
|
MC Fruit Stripe posted:Just fixed an issue which has plagued us since yesterday morning. Took a lot of time and got almost no sleep, but I think most of you guys understand how good it feels to fix something like that. King of the world baby. I had the same feeling a month ago when our HR Director quit.
|
|
#
?
Aug 3, 2017 22:47
|
|
|
I lol'd I've decided that when I join conference calls I want to be able to play Stone Cold Steve Austin's entrance music for the first 10 seconds or so. Need to figure out how to cue that up. A man needs an entrance.
|
|
#
?
Aug 3, 2017 22:54
|
|
|
MC Fruit Stripe posted:Just fixed an issue which has plagued us since yesterday morning. Took a lot of time and got almost no sleep, but I think most of you guys understand how good it feels to fix something like that. King of the world baby. you killed someone, didn't you
|
|
#
?
Aug 3, 2017 23:00
|
|
MC Fruit Stripe posted:Just fixed an issue which has plagued us since yesterday morning. Took a lot of time and got almost no sleep, but I think most of you guys understand how good it feels to fix something like that. King of the world baby. This happened to me a few months ago when read access was somehow removed from a directory partition resulting in a total loving of an entire domain. Still don't know how it happened 
|
|
|
#
?
Aug 3, 2017 23:21
|
|
|
|
| # ? May 25, 2024 21:42 |
|
|
CLAM DOWN posted:you killed someone, didn't you It's thinking outside of the box, but it resolved the issue.
|
|
#
?
Aug 3, 2017 23:34
|
|
