|
Yeah I saw that before, but the VM's need to restart with a host failure.. Can your SAN handle a bootstorm of that many VMs simultaneously without making GBS threads the bed? What I was getting at was with a bigger cluster of smaller specced hosts the impact of a host loss is much smaller, and it's easier to accommodate the failover capacity in the cluster.
|
#
?
Feb 8, 2013 04:01
|
|
|
|
| # ? May 8, 2024 22:47 |
|
|
What do you guys think is the sweet spot for power of a machine vs price? We bought Dell R710s that have two chips for around $10k each, but I would have felt a lot better buying single chip machines for $5k each. But Intel only has E5 chips in the latest generation that are for dual socket boards, so the best you could do is buy the R610 with two sockets and leave one empty for $7k.
|
|
#
?
Feb 8, 2013 04:18
|
|
|
GrandMaster posted:Who's gunna be the scrub when a host fails and takes out 700 VMs in one hit? FISHMANPET posted:What do you guys think is the sweet spot for power of a machine vs price? We bought Dell R710s that have two chips for around $10k each, but I would have felt a lot better buying single chip machines for $5k each. But Intel only has E5 chips in the latest generation that are for dual socket boards, so the best you could do is buy the R610 with two sockets and leave one empty for $7k. 
|
|
#
?
Feb 8, 2013 04:40
|
|
|
Crackbone posted:I work in a really small environment but I'm investigating virtualizing our entire setup w/VMWare. (As a side note, I'd love if somebody here wants to PM me for some general advice about the project and if it's even feasible). You kinda sound like you are doing what movax was doing a few pages back. What kind of servers and services are you looking at outside of your UTM? Martytoof posted:A little setup for my question: If OSX supports TRIM and is enabled to support it yes it will. Fusion is a Type 2 Hypervisor I believe, to the disk it will just look like a program is doing a bunch of I/O. GrandMaster posted:Yeah I saw that before, but the VM's need to restart with a host failure.. Can your SAN handle a bootstorm of that many VMs simultaneously without making GBS threads the bed? FISHMANPET posted:What do you guys think is the sweet spot for power of a machine vs price? We bought Dell R710s that have two chips for around $10k each, but I would have felt a lot better buying single chip machines for $5k each. But Intel only has E5 chips in the latest generation that are for dual socket boards, so the best you could do is buy the R610 with two sockets and leave one empty for $7k. This really depends on a number of factors such as; What is the VM services is the VM supporting the the company? How much resources is the VM taking up? What is the value of the Infrastructure the VM is supporting? What is the VM's estimated ROI? Scaling up hosts vs. Scaling out hosts is a tricky thing sometimes, as scaling out can often occur more than just the cost of a new server, there are a few things you have to take into effect such as; Cost per network uplinks used to connect the server, and if needed additional switching for the new host. Average host maintenance a year(BIOS, firmware, driver, and esxi updates) Power and Cooling costs of the additional host(s). And there are more factors to consider as well. Dilbert As FUCK fucked around with this message at 05:36 on Feb 8, 2013 |
|
#
?
Feb 8, 2013 05:09
|
|
|
Misogynist posted:Haha, x3850s don't fail, that's why they cost $75,000. You've clearly had a better IBM experience than me, all of our x3850/x3950s have been really unreliable. Mind you, these were a much older generation - from 7040 procs through to 51xx on our 3850's so maybe they got better. IBM also stopped certifying the hardware for ESX5+ when our dell kit from the same generation (PE2950's) are still good to go. Corvettefisher posted:EMC's FAST CACHE in the VNX's are a good example of how they can help reduce impacts of bootstorms Our new VNX5500 is due to land next thursday, with fast cache and fast vp. Building new datacenters is fun 
|
|
#
?
Feb 8, 2013 05:28
|
|
|
cheese-cube posted:What model is the blade? We have one 8853L5U, two 8853L6U, two 7995G6U, and a 8853PRL. The manuals say the 8853s all max out at 16 GB (4x4GB) (or 32 GB if we get an expansion blade) and the manuals for the 7995s say they max out at 32 GB (8x4GB)
|
|
#
?
Feb 8, 2013 05:30
|
|
|
GrandMaster posted:Our new VNX5500 is due to land next thursday, with fast cache and fast vp. Building new datacenters is fun Nice! The VNX's are really good for what you get, I am eager to see how some of the nimble stuff stacks against it.
|
|
#
?
Feb 8, 2013 05:40
|
|
|
For those of you who remember me posting questions over the past year about my weird NFS disconnects happening with my Netapp, well it looks like its finally officially a bug: http://virtualstorageguy.com/2013/02/08/heads-up-avoiding-vmware-vsphere-esxi-5-nfs-disconnect-issues/ For me it got solved when I upgraded my 2020 to a 2240 but I wish they would have figured this out sooner. Netapp pointed the finger at VMware, VMware pointed the finger at Netapp. I could have saved a ton of troubleshooting time if they figured this out sooner.
|
|
#
?
Feb 8, 2013 23:22
|
|
|
GrandMaster posted:I started building our new cluster yesterday - 16x Dell M620 2xE5-2670 / 256GB RAM
|
|
#
?
Feb 9, 2013 00:03
|
|
|
Yeah really, you just need (maybe) one vSwitch for management.
|
|
#
?
Feb 9, 2013 00:06
|
|
|
evil_bunnY posted:Yeah really, you just need (maybe) one vSwitch for management.
|
|
#
?
Feb 9, 2013 00:15
|
|
|
adorai posted:use a dvswitch, takes like 5 minutes for us to configure networking on a host. No money left in the budget for the uplift to ent+ licensing 
|
|
#
?
Feb 9, 2013 00:21
|
|
|
GrandMaster posted:You've clearly had a better IBM experience than me, all of our x3850/x3950s have been really unreliable. Mind you, these were a much older generation - from 7040 procs through to 51xx on our 3850's so maybe they got better. IBM also stopped certifying the hardware for ESX5+ when our dell kit from the same generation (PE2950's) are still good to go. I took over the IT dept at my company a few years ago after my then boss quit and moved onto another company. Turns out he hadn't updated the firmwares of anything in forever and we were on VMware 3.5. During the process of getting everything up to spec prior to moving to 5.0u1, IBM managed to fry the UEFI on two of our servers because they didn't provide the proper firmware updates. We were so far behind that they should have had us hop to a certain level before attempting to get us to the latest version. The 2nd time this happened, I told them what happened the first time and questioned if we needed to hop to a previous version before attempting yet they assured me it would work. Another time, a hard drive failed on a RAID-5 array at our satellite office that only has one host and no SAN. An apparent firmware bug caused the wrong drive to illuminate its error light and subsequently, I yanked a good drive out of a RAID-5 array when another drive was already down. IBM ended up throwing their hands up in the air and told me to recover from backup. I was able to rebuild the array no thanks to them. Also, when you're paying thousands of dollars for a support contract annually, not being able to be connected to a technician immediately for a production server down incident is loving asinine. My experiences with VMware's support is the polar opposite. I have nothing but good things to say about their support team.
|
|
#
?
Feb 9, 2013 00:28
|
|
|
goobernoodles posted:Also, when you're paying thousands of dollars for a support contract annually, not being able to be connected to a technician immediately for a production server down incident is loving asinine. IBM does this -- without hassle or even a question -- but they don't readily disclose the process. Open up a ticket using the online ESC+ tool. As soon as you have your confirmation that the ticket was submitted and you have a case number, call 1-800-IBM-SERV and get somebody, anybody, on the phone. Once you have a warm body, the second they start to ask you for your personal information, ask to speak with the National Duty Manager. You can escalate your case immediately from there. If you want, they'll often stay on the call with you and the technician until they're satisfied you have the resolution you're looking for. Our issue is that one of the engineers in my group is hearing-impaired, which obviously makes this telephone-based route rather difficult. The National Duty Manager I spoke to about this had no idea herself how someone with a hearing impairment is supposed to escalate a case. IBM's support is really rather good from a personnel and policy standpoint, but their processes are dumb. goobernoodles posted:I took over the IT dept at my company a few years ago after my then boss quit and moved onto another company. Turns out he hadn't updated the firmwares of anything in forever and we were on VMware 3.5. During the process of getting everything up to spec prior to moving to 5.0u1, IBM managed to fry the UEFI on two of our servers because they didn't provide the proper firmware updates. We were so far behind that they should have had us hop to a certain level before attempting to get us to the latest version. The 2nd time this happened, I told them what happened the first time and questioned if we needed to hop to a previous version before attempting yet they assured me it would work. I'm really surprised you hit this problem on M2+ servers, though, at least in a way serious enough to impact production. We've botched firmware upgrades and been able to restore to the factory UEFI image automatically. Vulture Culture fucked around with this message at 07:53 on Feb 9, 2013 |
|
#
?
Feb 9, 2013 07:42
|
|
|
Anyone knows whether HyperV supports TRIM on their VHDs, e.g. to compact them when using them with a not-Windows OS? And if so, I suppose any writes done without the synthetic drivers won't be affected?
|
|
#
?
Feb 10, 2013 17:14
|
|
|
loving great the vSGA demo got cancelled for PEX.... I am pretty pissed.
|
|
#
?
Feb 11, 2013 18:42
|
|
|
So, has anyone gone through and implemented SSL certs on an existing environment that was setup using self signed? If so, any kind of "gotchas" going along with this? I just discovered that our devs setup a CA setup. I am hoping to assign certs to our vCenter server as well as on each host.
|
|
#
?
Feb 11, 2013 19:03
|
|
|
Not really just follow the Whitepaper and it should be fairly straight forward.
|
|
#
?
Feb 11, 2013 19:10
|
|
|
Corvettefisher posted:Not really just follow the Whitepaper and it should be fairly straight forward. Thanks. Just wanted to check to see if there was anything I was blatantly missing.
|
|
#
?
Feb 11, 2013 19:21
|
|
|
Moey posted:Thanks. Just wanted to check to see if there was anything I was blatantly missing. I came across a script for this a while back, may or may not be worth the time vs. doing it manually depending on how many hosts you're talking about.
|
|
#
?
Feb 11, 2013 19:24
|
|
|
I have old phone server software that I installed on a Windows 2000 Server VM using Hyper-V on Windows Server 2008 R2. The software has a hardware key that plugs into a parallel port. We forwarded the port to the VM, however the VM is not recognizing the hardware key. I noticed that one workaround that people are using for USB hardware keys is to plug them into an Network-Attached USB hub. Do you think that I could get away with plugging a USB to parallel adapter into a Network-Attached USB hub or is there an easier way to get the parallel port hardware key working? I do have a parallel card in the server. Also, would running Hyper-V as an administrator make any difference? Does VMWare have better support for parallel ports? I'm not too experienced with virtual machines, so sorry if some my questions are a bit basic. Thank you in advance for any help.
|
|
#
?
Feb 11, 2013 23:20
|
|
|
Aniki posted:I have old phone server software that I installed on a Windows 2000 Server VM using Hyper-V on Windows Server 2008 R2. The software has a hardware key that plugs into a parallel port. We forwarded the port to the VM, however the VM is not recognizing the hardware key. I noticed that one workaround that people are using for USB hardware keys is to plug them into an Network-Attached USB hub. Do you think that I could get away with plugging a USB to parallel adapter into a Network-Attached USB hub or is there an easier way to get the parallel port hardware key working? I do have a parallel card in the server. Also, would running Hyper-V as an administrator make any difference? Does VMWare have better support for parallel ports?
|
|
#
?
Feb 12, 2013 01:19
|
|
|
Quite probably a dumb question, but I've finally given up trying to figure out what I've missed. I P2V'd my work laptop using converter and dropped the file onto my home computer. I booted up w/ VMware Workstation just fine into Win7 once. I removed what physical devices I thought I saw and then rebooted. I cannot get back into normal mode but I can get into safe mode. I lock hard while booting into Win7, have to Reset power to get it back. I feel like such a dork, I can't figure out how to troubleshoot windows not booting. I've quite literally have never had this issue and am stumped. I'd like a log file or something to indicate where the heck it's hanging. If this doesn't work, it's just an exercise. I wanted to P2V my laptop so I could run the VPN from within workstation and leave my physical laptop at work. They only assign us one RSA soft token, and it's on the laptop. Would've been nice to run my 'laptop' from within my home PC and work from home over the VPN. I'm not entirely sure the RSA token would work from within a VM, I'm not keen on how RSA soft tokens authenticate to the 'host' it's on; if at all. SecureID is witchcraft for all I know. An Escalation Engineer at my office said he's done the converter process. I's possible he's loving with me just to watch me flounder, but he seemed sincere.
|
|
#
?
Feb 12, 2013 06:20
|
|
|
It's been a while, as we stopped using our RSA system, but I think if you can get the original seed file for your token you can install the soft token in multiple places. I recall having the softtoken on my laptop and my phone and either one could be used.
|
|
#
?
Feb 12, 2013 06:39
|
|
|
talaena posted:Quite probably a dumb question, but I've finally given up trying to figure out what I've missed. I P2V'd my work laptop using converter and dropped the file onto my home computer. I booted up w/ VMware Workstation just fine into Win7 once. I removed what physical devices I thought I saw and then rebooted. I cannot get back into normal mode but I can get into safe mode. I lock hard while booting into Win7, have to Reset power to get it back. I'm not an expert by any means, but the conversion process should remove any "physical" hardware at the point the conversion is done - or at least anything detrimental to the operation of the machine. If it booted up fine after the initial you probably didn't need to remove anything from device manager. If you were told something different I'd certainly trust your tech more though.
|
|
#
?
Feb 12, 2013 14:18
|
|
|
skipdogg posted:It's been a while, as we stopped using our RSA system, but I think if you can get the original seed file for your token you can install the soft token in multiple places. I recall having the softtoken on my laptop and my phone and either one could be used. The distributed token file may optionally contain information to lock it to a single host (I believe this is based on the serial of the HD the token store resides on for Win/OSX tokens, unless you're using some other storage like a USB token). If your org is using this they would have asked you for the DeviceSerialNumber (which you can find in the software token application under Options > Token Storage Devices) when you requested the token.
|
|
#
?
Feb 12, 2013 15:52
|
|
|
talaena posted:Quite probably a dumb question, but I've finally given up trying to figure out what I've missed. I P2V'd my work laptop using converter and dropped the file onto my home computer. I booted up w/ VMware Workstation just fine into Win7 once. I removed what physical devices I thought I saw and then rebooted. I cannot get back into normal mode but I can get into safe mode. I lock hard while booting into Win7, have to Reset power to get it back. What devices did you remove following the initial successful boot?
|
|
#
?
Feb 12, 2013 15:58
|
|
|
adorai posted:In my experience getting hardware licensing HASPs to work with a USB to parrallel adapter is hit or miss, but if it works with the cable, it should be trivial to use the USB to LAN hub. We do it and other than it occassionally needing to be reconnected, it's been pretty good. Ok, I'll try plugging in a USB to parallel adapter to the server and see if I can get that working tonight, though it is good to know that the USB to LAN hub likely won't make a difference.
|
|
#
?
Feb 12, 2013 17:39
|
|
|
It can be really hit or miss whether the USB2LAN device works with your particular token though. And DR is still annoying to plan for.
|
|
#
?
Feb 12, 2013 17:51
|
|
|
evil_bunnY posted:It can be really hit or miss whether the USB2LAN device works with your particular token though. And DR is still annoying to plan for. I think my plan is to try the USB to Parallel adapter first and if that doesn't work, then I'll consider trying VMWare (currently using Hyper-V) or ordering the USB2LAN hub. The hardware key is for Call Cener Worx v. 2.1, which was released in 2001 and can only run on Windows NT based operating systems and for some reason they stopped purchasing updates after that. I know that the USB keys they released later on were finicky and I'm not sure if the parallel key we have is supposed to be any better.
|
|
#
?
Feb 12, 2013 19:43
|
|
|
Does anyone here run KVM in a serious business environment?
|
|
#
?
Feb 14, 2013 04:36
|
|
|
adorai posted:Does anyone here run KVM in a serious business environment? Guest OSes are all Redhat or Centos 5 and 6, Ubuntu 12.04 and Windows Server 2008r2. So far KVM has been really nice and rock solid for us, no weird guest crashes or weirdness that we've seen. Did you have any specific questions, or were you just wondering if anyone does it?
|
|
#
?
Feb 14, 2013 05:12
|
|
|
Those of you running giant VMs: how is co-scheduling in ESXi 5.1? I have a Puppet server I'd really like to scale up before scaling out, but I'm debating whether trying to go 6 vCPUs on a VM will help or hurt here. The hosts are 2×8-core Sandy Bridge servers.
|
|
#
?
Feb 14, 2013 06:06
|
|
|
adorai posted:Does anyone here run KVM in a serious business environment? RHEV where I work. Most of the companies that use OpenStack use KVM as their hypervisor too.
|
|
#
?
Feb 14, 2013 07:07
|
|
|
Less Fat Luke posted:Did you have any specific questions, or were you just wondering if anyone does it?
|
|
#
?
Feb 14, 2013 13:37
|
|
|
Misogynist posted:Those of you running giant VMs: how is co-scheduling in ESXi 5.1? I have a Puppet server I'd really like to scale up before scaling out, but I'm debating whether trying to go 6 vCPUs on a VM will help or hurt here. The hosts are 2×8-core Sandy Bridge servers. Depends on how saturated the box you are putting this on is already. I don't really have any problem with some of the 4vcpu VM's on my E5649 6c/12t Westmere CPU's, the box isn't incredibly saturated however, it is only running about 12 or so non high load VM's. Haven't tested 6 though. On a different topic, does anyone else LOVE view 5.1.2 as much as I do? I sure do love the Tshooting, most things end up being "Well just rebuild the service/module/environment". Thanks for rushing View 5.1.x VMware, really appreciate the lack of QA.
|
|
#
?
Feb 14, 2013 18:56
|
|
|
adorai posted:really just looking for the warm and fuzzy that someone actually does it.
|
|
#
?
Feb 14, 2013 19:15
|
|
|
Less Fat Luke posted:Yeah I'd vouch for it anytime. The only thing in our setup that gave us grief was rolling our own clustered NFS with DRBD. It works fine now, but there were a lot of loving annoying stumbling blocks along the way to get it there and in the end it's probably less time and grief just to buy commercial network storage. Do you have any tips on working with DRBD? I'm basically implementing a replicating SAN in linux using DRBD since our applications aren't allowed to have RDMs presented to them for storage. It's stupid but whatever. I'm just hoping there aren't any obvious gotchas.
|
|
#
?
Feb 14, 2013 19:45
|
|
|
Goon Matchmaker posted:Do you have any tips on working with DRBD? I'm basically implementing a replicating SAN in linux using DRBD since our applications aren't allowed to have RDMs presented to them for storage. It's stupid but whatever. I'm just hoping there aren't any obvious gotchas.
|
|
#
?
Feb 14, 2013 21:24
|
|
|
|
| # ? May 8, 2024 22:47 |
|
|
Misogynist posted:DRBD is incredibly fragile and breaks all the time and you're going to want to kill yourself. Other than this, have a great time! Don't tell me this Edit: Just found lsyncd. This suits our needs way better than DRBD and should be more reliable. Goon Matchmaker fucked around with this message at 21:42 on Feb 14, 2013 |
|
#
?
Feb 14, 2013 21:25
|
|