|
Richard Noggin posted:For all of you budding server admins: when you're provisioning storage, for the love of Christ please don't carve out a 2TB volume when all you need is 30GB. If you need more space down the road, it's really easy to extend volumes in Windows (and probably Linux, but that's not my area) but quite the pain in the rear end to shrink them to a more manageable size after the fact. But storage is cheap 
|
#
?
Jan 13, 2015 17:43
|
|
|
|
| # ? Apr 25, 2024 03:22 |
|
|
BigPaddy posted:But storage is cheap So are replacement retards.
|
|
#
?
Jan 13, 2015 17:44
|
|
|
Richard Noggin posted:For all of you budding server admins: when you're provisioning storage, for the love of Christ please don't carve out a 2TB volume when all you need is 30GB. If you need more space down the road, it's really easy to extend volumes in Windows (and probably Linux, but that's not my area) but quite the pain in the rear end to shrink them to a more manageable size after the fact. Conversely, If you're administering and provisioning storage for Christ's sake allow for a sane system drive size! Extending system drives is scary and something to avoid! I have these companies that have a standard of 16-20GB OS Drives. Seriously! Thin provision me but at LEAST give me 30GB so the drat thing doesn't run out of space down the line due to patches.
|
|
#
?
Jan 13, 2015 17:50
|
|
|
Zaepho posted:Conversely, If you're administering and provisioning storage for Christ's sake allow for a sane system drive size! Extending system drives is scary and something to avoid! I have these companies that have a standard of 16-20GB OS Drives. Seriously! Thin provision me but at LEAST give me 30GB so the drat thing doesn't run out of space down the line due to patches. Extending system drives on Server 2008 and up is trivial. I've done it countless times and never had a single problem. That's not to say you shouldn't have a recent backup to CYA, but there's really nothing to be scared of.
|
|
#
?
Jan 13, 2015 17:55
|
|
|
Remember the good old days when that was the loving factory default from Dell? I haven't bought a server with Windows preinstalled in a very long time but I assume that's not the case anymore (right?  ). But in the Server 2003 era, god drat, so many servers with like a 10GB system drive and gigantic D: drive.
|
|
#
?
Jan 13, 2015 18:00
|
|
|
BigPaddy posted:I used too before the current bunch I worked with found an Agile Development book somewhere and decided that telling me to develop everything in an Agile way means they tell me nothing and I have to guess what they want and when it isn't have to listen to their shrill ear splitting complaining that I didn't read their mind well enough. Agile can be great if people are bought in. It sounds like you aren't, because you want big design up front instead of to work with the users to figure out what they actually want. (This is an assumption on my part, because you wouldn't have to guess if you would be a human and ask questions instead.) Read Jeff Patton's User Story Mapping book for a good perspective on how to do this. Vulture Culture fucked around with this message at 18:10 on Jan 13, 2015 |
|
#
?
Jan 13, 2015 18:06
|
|
|
Zaepho posted:Conversely, If you're administering and provisioning storage for Christ's sake allow for a sane system drive size! Extending system drives is scary and something to avoid! I have these companies that have a standard of 16-20GB OS Drives. Seriously! Thin provision me but at LEAST give me 30GB so the drat thing doesn't run out of space down the line due to patches. Extending drives on Server 2008 R2 and newer is a speedy and smooth non-issue and not even remotely scary or something to avoid. If it is, you're doing something wrong! e:fb
|
|
#
?
Jan 13, 2015 18:06
|
|
|
Richard Noggin posted:Extending system drives on Server 2008 and up is trivial. I've done it countless times and never had a single problem. That's not to say you shouldn't have a recent backup to CYA, but there's really nothing to be scared of. It's still so much easier to recover data than the OS that I'd rather avoid it. Luckily several horrible experiences in the distant past have made me pretty adamant about a decent system drive size so I haven't had to mess with it since the Server 2000/2003 days. Glad to hear it's better these days! I'll still avoid it though...
|
|
#
?
Jan 13, 2015 18:12
|
|
|
Misogynist posted:I think it's good for business units to have these conversations with each other, provided that the conversations are well-informed. Doing this is pretty much the sole function of any IT department that treats itself as a competitive advantage rather than a cost center. Hey, that process you have sucks! Maybe you should automate it! Let me show you how! Likewise, IT people tend to be pretty loving awful at managing and executing projects. Agreed, IT need to be involved and proactive. The problem is that when you are in an environment where everyone views you as a cost centre and is generally not open to IT feeding ideas on business process improvement because "You are just the IT guy" it makes it hard to work with them when you are being decreed the system will work like this without any feedback loop between IT and the business being in place. This is a cultural issue that comes all the way from the CEO who will issue his decrees and expect things to be done and then IT get pushed into trying to delivery something in the timeframe. At the moment we have 9 major projects running each of them allegedly vital and IT people are pulled between these on a near weekly basis depending on which project manager has cried the loudest in the last week. Timelines, scope and cost are dictated without any planning being done which means every project is late, over budget and buggy which leads to every project then needing a second project to fix the first which again is planned without any realistic timeline, cost or scope so it suffers the same fate and push back on the fact that all our projects are late, over budget and buggy is pushed aside. On top of this there is a revolving door of external consultancy companies that business units hire and don't say anything or have IT deal with before any contract has been sign and the first thing you know about them is when you get in Monday morning and see meeting invites for the whole week to have a workshop about a project you know nothing about. Misogynist posted:Agile can be great if people are bought in. It sounds like you aren't, because you want big design up front instead of to work with the users to figure out what they actually want. (This is an assumption on my part, because you wouldn't have to guess if you would be a human and ask questions instead.) Read Jeff Patton's User Story Mapping book for a good perspective on how to do this. Agile is fine if the business want to give very top level requirements but then they can't complain to the CTO and Director of IT when they get something at the end of the first few sprints demo'd to them that isn't the perfect system they envisioned. The people I work with are not invested enough into Agile to understand yes you get stuff quicker but it might not be what you wanted straight away. Also the business units do not talk to each other. For example Marketing and Sales last year both paid external companies to do data clean up in our CRM systems. When I pointed out to each other that they were both paying for the same thing and maybe they should talk to each other the response was very much it was their budget and they will spend on what they like. So yes considering the people I have to work with I want more details up front so I don't have to have those meetings with my boss and his boss about why this latest project didn't give Business Unit X what they wanted straight away and why is it late and why did you go over budget. User Story Mapping and all that fun stuff is great when you have business units who want to work with you rather than business units who want to shoot you a two line email of what they wanted yesterday and for 10 quid.
|
|
#
?
Jan 13, 2015 18:22
|
|
|
Docjowles posted:Remember the good old days when that was the loving factory default from Dell? I haven't bought a server with Windows preinstalled in a very long time but I assume that's not the case anymore (right? And here I thought it was the morons who worked here in the past that left me with the 10GB C: volumes on my now virtual 2003 servers. Not saying they are in the clear, but gently caress Dell for doing that and causing me all kinds of pain because SEP packages are like 400mb and our system doesn't clean up after itself.
|
|
#
?
Jan 13, 2015 18:50
|
|
|
To be fair, they are still morons for not nuking that poo poo and creating a more reasonable size. We have a few here that were like that and I've ended up having to resize them so that we could get them patched not poo poo the bed every time new patches are released.
|
|
#
?
Jan 13, 2015 18:58
|
|
|
poo poo, my default OS drive is 60GB for my server builds. 2008R2 will drat near use 35 after install and patching.
|
|
#
?
Jan 13, 2015 19:08
|
|
|
skipdogg posted:poo poo, my default OS drive is 60GB for my server builds. 2008R2 will drat near use 35 after install and patching. What the heck man, it does not. 2008R2, on a fresh install, after all updates and a couple gigs of our corporate baseline applications, is using 22.4GB on the OS disk, looking at it right now.
|
|
#
?
Jan 13, 2015 19:34
|
|
|
skipdogg posted:poo poo, my default OS drive is 60GB for my server builds. 2008R2 will drat near use 35 after install and patching. I usually feel like 60 is a very reasonable number. Enough to deal with a couple years of patches without making GBS threads itself. plus the occasional larger profile or app that just HAS to install binaries to c:\Program Files\StupidVendor\ or it explodes it a fit of shame and hilarity.
|
|
#
?
Jan 13, 2015 19:36
|
|
|
I honestly had no idea that some people actually used any OEM machines as configured out-of-the-box. 
|
|
#
?
Jan 13, 2015 19:38
|
|
|
Japanese Dating Sim posted:I honestly had no idea that some people actually used any OEM machines as configured out-of-the-box. Most of the ones I encountered were at small businesses where poo poo had been set up by whatever random employee was "good with computers". They weren't actually good with computers
|
|
#
?
Jan 13, 2015 19:46
|
|
|
So, I've been working at an entry-level position here for the past few months. I posted about it here before. I am totally enjoying it, especially since I am given plenty of opportunity to do things beyond answering help desk calls. Anyways, I was working on configuring some Meraki switches for the past few hours. I had them cabled correctly into a switch that had internet access, but was dumbfounded as to why they were not able to access the internet themselves. Tried all the troubleshooting stuff I could think of. Turns out, the original switch and the network's addressing scheme is 172.19.2.0 instead of the 172.16.2.0 I saw when I looked at it the first time. So, the only problem was that six instead of a nine. Figured you people would get a laugh out of that.
|
|
#
?
Jan 13, 2015 20:08
|
|
|
CLAM DOWN posted:What the heck man, it does not. 2008R2, on a fresh install, after all updates and a couple gigs of our corporate baseline applications, is using 22.4GB on the OS disk, looking at it right now. I have a brand new 2008 install that is 35 gigs after service pack, powershell, and two days of windows updates.
|
|
#
?
Jan 13, 2015 20:48
|
|
|
thebigcow posted:I have a brand new 2008 install that is 35 gigs after service pack, powershell, and two days of windows updates. 2008 or 2008R2? And I'm not sure how that's possible, what you're saying doesn't make sense for a 2008R2 machine. PowerShell 2.0 is installed by default, if you mean 4.0 that's an 18MB patch for the WMF update, and the initial run of Windows updates on a fresh 2008R2 build takes a couple hours at most, not 2 days. And I can screenshot the OS disk usage if you'd like, 22.4GB.
|
|
#
?
Jan 13, 2015 21:02
|
|
|
I know it's not apples, but I have a two day old install of 2012 R2 consuming 17.5GB.
|
|
#
?
Jan 13, 2015 21:10
|
|
|
CLAM DOWN posted:2008 or 2008R2? And I'm not sure how that's possible, what you're saying doesn't make sense for a 2008R2 machine. PowerShell 2.0 is installed by default, if you mean 4.0 that's an 18MB patch for the WMF update, and the initial run of Windows updates on a fresh 2008R2 build takes a couple hours at most, not 2 days. And I can screenshot the OS disk usage if you'd like, 22.4GB. 2008
|
|
#
?
Jan 13, 2015 22:14
|
|
|
CLAM DOWN posted:2008 or 2008R2? And I'm not sure how that's possible, what you're saying doesn't make sense for a 2008R2 machine. PowerShell 2.0 is installed by default, if you mean 4.0 that's an 18MB patch for the WMF update, and the initial run of Windows updates on a fresh 2008R2 build takes a couple hours at most, not 2 days. And I can screenshot the OS disk usage if you'd like, 22.4GB. 2008 R2 updates, if applied over a long enough time, can bloat WinSxS. One of the last major things microsoft did to 2008 R2 was to allow cleanup. However its not without its downsides: It requires the Windows Desktop experience installed. Not ideal for hardened deployments.
|
|
#
?
Jan 13, 2015 23:37
|
|
|
Japanese Dating Sim posted:I honestly had no idea that some people actually used any OEM machines as configured out-of-the-box. I created an MDT Post-OS task sequence system with scripts to remove HP/Dell bloatware after getting sick of setting up Windows OEM desktops... it works really well. I work for an MSP supporting loads of small businesses so "just make an image" wasn't really an option for a variety of reasons.
|
|
#
?
Jan 13, 2015 23:44
|
|
|
dox posted:I created an MDT Post-OS task sequence system with scripts to remove HP/Dell bloatware after getting sick of setting up Windows OEM desktops... it works really well. I work for an MSP supporting loads of small businesses so "just make an image" wasn't really an option for a variety of reasons. I should really do this.
|
|
#
?
Jan 13, 2015 23:51
|
|
|
mayodreams posted:We have it in production as part of our order management system but I really don't touch it. I've played with the free version a few years ago and it was a pain to get up an running. Would you use this or recommend to manage, oh I don't know.......30 million documents (14TB or so)?
|
|
#
?
Jan 13, 2015 23:55
|
|
|
I have what I'm pretty sure is a really dumb question and I'll just throw it out there - there's no way to get around needing a pair of human eyes to look at MAC addresses on new machines for whitelisting purposes, right? I can't think of how you'd get around that but I'm still newish to IT so  .Edit: dox posted:I created an MDT Post-OS task sequence system with scripts to remove HP/Dell bloatware after getting sick of setting up Windows OEM desktops... it works really well. I work for an MSP supporting loads of small businesses so "just make an image" wasn't really an option for a variety of reasons.
|
|
#
?
Jan 13, 2015 23:59
|
|
|
Would conducting an ARP scan give you what you're looking for? I'm even newer than you and probably sound dumb trying to answer that question but I'm curious about it myself!
|
|
#
?
Jan 14, 2015 00:29
|
|
|
incoherent posted:Would you use this or recommend to manage, oh I don't know.......30 million documents (14TB or so)? I'd probably be happier with that than Sharepoint, Xerox's DocuShare, or yuck: Jive.
|
|
#
?
Jan 14, 2015 01:31
|
|
|
Roargasm posted:I'm green and probably have a lighter workload than you, but I started going gray for a couple of months and was loving losing it and I read Limoncelli's Time Management for System Administrators (there's a Kindle edition). He dives right into the core concept that having to remember 500 things at once (my job, and I assume yours) ruins your alacrity on the job, even if you don't consciously realize it. I started writing down absolutely everything and only focusing on what was right in front of me. My stress level went way down, I stopped worrying about fires and focused on the work I was doing, which hopefully leads to fewer fires anyway. My time management is a big problem, thanks for the book recommendation it should help a great deal! The main issue I have is that it's myself and one other guy who can handle the infrastructure/linux and we have about 30 sans, 5 full blade chassis and a shitload of ageing hardware still in production and after a year of terrible firmware issues on my sans, potential disasters are always on my mind. I've also gone from an 8 week pager rotation to a 2 week because people have left and we haven't hired for some reason. Really the solution involves getting another job I think, away from hosting. Japanese Dating Sim posted:I have what I'm pretty sure is a really dumb question and I'll just throw it out there - there's no way to get around needing a pair of human eyes to look at MAC addresses on new machines for whitelisting purposes, right? I can't think of how you'd get around that but I'm still newish to IT so I know some vendors ship their computers with a MAC address barcode on the box. barcode scanners are cheap. Dealing with stupid mac address whitelists is a pain in the rear end, so happy I don't have that here any more.
|
|
#
?
Jan 14, 2015 02:30
|
|
|
Japanese Dating Sim posted:I have what I'm pretty sure is a really dumb question and I'll just throw it out there - there's no way to get around needing a pair of human eyes to look at MAC addresses on new machines for whitelisting purposes, right? I can't think of how you'd get around that but I'm still newish to IT so
|
|
#
?
Jan 14, 2015 02:40
|
|
|
theperminator posted:My time management is a big problem, thanks for the book recommendation it should help a great deal!
|
|
#
?
Jan 14, 2015 02:44
|
|
|
Yeah, it's a pretty large virtual environment.
|
|
#
?
Jan 14, 2015 02:55
|
|
|
theperminator posted:Yeah, it's a pretty large virtual environment.
|
|
#
?
Jan 14, 2015 03:03
|
|
|
Richard Noggin posted:For all of you budding server admins: when you're provisioning storage, for the love of Christ please don't carve out a 2TB volume when all you need is 30GB. If you need more space down the road, it's really easy to extend volumes in Windows (and probably Linux, but that's not my area) but quite the pain in the rear end to shrink them to a more manageable size after the fact. Last week our vCenter server's C:\ drive needed to be extended (it's virtualized), so they went ahead and did it, but something went wrong and caused it to crash and bring down our entire virtual infrastructure. For most places that would be horrifying and a Resume Generating Event for the person who did it, but we managed to bring ALL email for the USAF down and, according to quite a few flag officers, seriously endanger national security. It took a few hours to get everything back up and running, and they were still cleaning things up over the weekend. Since I was off until Saturday evening I didn't even hear about it until someone started talking about how much of a bitch it was, and how the phones just blew up. Not sure if we have an opening on our Virtualization team now - I should probably start asking around. Of course, last night into this morning things were also a pain - someone at INOSC East decided to gently caress around with the routers and prevented people across the AF who were off-base from being able to access email through OWA or connect via VPN. We kept telling ESD over and over that it was a Boundary issue and that email was working just fine, but they still kept giving users our phone number to call, and tickets that ESD created and sent to us were forwarded to Boundary, who didn't even look at them before bouncing them back. INOSC East Crew Commanders were totally unimpressed with our argument that if people who were on-base were able to get their email, and people who were off-base couldn't, then just maybe the problem was with the interface (i.e. routers), which they just so conveniently happened to control and, also conveniently, happened to have made changes to. I'm curious to see what happened after I left this morning. As a side note, rumor has it that our contract is going to get extended for a year because the organizational changes made didn't quite produce the results the USAF had been expecting, and our detachment is mostly the reason why things haven't completely degenerated into mass chaos. They still want to kill us off, but the officers also want their promotions, so they're thinking about giving us a stay of execution (they've been trying to get rid of the detachment since 2006, but each time the hammer is about to fall someone gets a dose of common sense).
|
|
#
?
Jan 14, 2015 03:08
|
|
|
Daylen Drazzi posted:Last week our vCenter server's C:\ drive needed to be extended (it's virtualized), so they went ahead and did it, but something went wrong and caused it to crash and bring down our entire virtual infrastructure. For most places that would be horrifying and a Resume Generating Event for the person who did it, but we managed to bring ALL email for the USAF down and, according to quite a few flag officers, seriously endanger national security. It took a few hours to get everything back up and running, and they were still cleaning things up over the weekend. Since I was off until Saturday evening I didn't even hear about it until someone started talking about how much of a bitch it was, and how the phones just blew up. Not sure if we have an opening on our Virtualization team now - I should probably start asking around. I don't understand how a vcenter server could bring down the entire infrastructure. Are you sure it was the vcenter server?
|
|
#
?
Jan 14, 2015 03:40
|
|
|
If you lose vcenter, don't you also lose SSO and HA?
|
|
#
?
Jan 14, 2015 03:44
|
|
|
Tab8715 posted:If you lose vcenter, don't you also lose SSO and HA? Unless I am really hosed up in my understanding of vcenter, no it shouldn't. The vmware HA agents on each host in the cluster wouldn't do anything by seeing a vcenter server go offline and would work how they were last configured in HA situations. I have never heard of a situation where you take your vcenter server offline (patching, reboots, whatever) and fear your all your vm's are suddenly not going to work. Sickening fucked around with this message at 03:53 on Jan 14, 2015 |
|
#
?
Jan 14, 2015 03:50
|
|
|
adorai posted:I'm curious, what are you whitelisting mac addresses on? What's going on here? I work in a university that uses Infoblox. Devices have to have their MAC addresses registered to receive an IP address from DHCP, so part of the process for a new machine is that we look up the MAC address and add them in prior to imaging them. Is this unusual? I kinda thought most places had a whitelist structure, but again, pretty new.
|
|
#
?
Jan 14, 2015 04:02
|
|
|
Japanese Dating Sim posted:I work in a university that uses Infoblox. Devices have to have their MAC addresses registered to receive an IP address from DHCP, so part of the process for a new machine is that we look up the MAC address and add them in prior to imaging them. Although I somewhat admire the security put in place that has to be extremely inconvenient and someone is being a little lazy in your administration.
|
|
#
?
Jan 14, 2015 04:05
|
|
|
|
| # ? Apr 25, 2024 03:22 |
|
|
Whitelisting is pretty much non-existent due to overhead and management constraints. But then again, it's in healthcare so who knows. I'm long past being surprised at what that sector gets up to. Something that's a little sideways, but if these are physical connections onto a managed switch you can use something like Cisco's CDP to get mac addresses on specific ports. Even if it's not Cisco, if you can just plug it into a specific network port, you'll be able to read / pull the mac address without needing a DHCP lease. You just might have to log into the switch to get the info, but all the switch needs is link up to read - doesn't need an OS.
|
|
#
?
Jan 14, 2015 04:12
|
|
