|
the spyder posted:My sales rep just tried to sell me vSphere standard edition... For my two, 192gb ram, dual 8 core hosts... He claimed it supported HA and vMotion, and that I only needed one license... Do not listen to this guy. 192gb in each host, or total?
|
# ? Jun 6, 2012 22:45 |
|
|
# ? May 15, 2024 04:14 |
|
DrOgdenWernstrom posted:vSphere standard is by CPU, so I believe you would need four licenses of vSphere standard to cover the cpu's, but it has a 32gb/license limit on RAM.
|
# ? Jun 6, 2012 22:47 |
|
Misogynist posted:He can have 192 GB in one server if that's all he wants to use, vRAM entitlements are cumulative and not averaged across your environment. I could be reading it wrong, but it sounds like he may have 192gb mem per server, which would make licensing 5 pricy.
|
# ? Jun 6, 2012 22:52 |
|
Erwin posted:You really give every VM 2 vCPUs, or am I reading that wrong? No, I meant as a minimum generally, I didn't mean all VMs get a fixed number or something
|
# ? Jun 6, 2012 23:25 |
|
I was curious if any of you knew of any "right-size" tools for Xenserver? I love Veeam One as it'll give me an extremely simple breakdown to bring to management. I'll be moving off of this Xenserver environment shortly (once I've received my Vsphere and new Veeam licenses) but I'd like to get a good idea of the current environments ACTUAL resource needs. I inherited this cluster, and a huge number of servers are just hanging out mostly idle and greatly overprovisioned- but I'd like an automated tool to report that to me over a time period.
|
# ? Jun 6, 2012 23:33 |
|
HalloKitty posted:No, I meant as a minimum generally, I didn't mean all VMs get a fixed number or something A minimum of 2 is what I thought you meant. That's really stupid.
|
# ? Jun 6, 2012 23:37 |
|
Misogynist posted:He can have 192 GB in one server if that's all he wants to use, vRAM entitlements are cumulative and not averaged across your environment. http://www.vmware.com/products/datacenter-virtualization/vsphere/small-business/compare-kits.html Under vRam entitlement for Essentials Plus, why does it say "32GB (192GB total)" if it's cumulative?
|
# ? Jun 6, 2012 23:40 |
|
HalloKitty posted:No, I meant as a minimum generally, I didn't mean all VMs get a fixed number or something I would strongly suggest you read up on coscheduling. Allocating two vCPUs when it is not needed will typically reduce performance not just overall, but also on that specific VM.
|
# ? Jun 6, 2012 23:55 |
|
got it working. Its drinking time.
|
# ? Jun 7, 2012 00:02 |
|
DrOgdenWernstrom posted:http://www.vmware.com/products/datacenter-virtualization/vsphere/small-business/compare-kits.html I have 192gb total across two boxes. However I hope it is not 32gb per machine, but per CPU, which still would not make since. This is a good question and it looks like the Essentials Kit is what I am after. But first I need my E5-2650's to arrive.
|
# ? Jun 7, 2012 01:04 |
|
The "total 192GB" is code:
It's definitely not 32GB per host as you feared; the standard setup for Essentials Plus would be 3 hosts, 2 pCPUs/host, 64GB pRAM/host. I think the Essentials Plus licenses might come as 6 CPU licenses you apply to the ESXi hosts directly, so whether or not you can apply 3 licenses on each of 2 hosts (as opposed to the "intended" 2 licenses on each of 3 hosts) is a good question.
|
# ? Jun 7, 2012 01:18 |
|
HalloKitty posted:No, I meant as a minimum generally, I didn't mean all VMs get a fixed number or something Chapter 6. I don't know a whole lot about VMware yet, but I do know that you're wrong.
|
# ? Jun 7, 2012 05:07 |
|
It's good to know that before I inadvertently spread any other misinformation. Always good to find out the right way to do things, no problem admitting I'm wrong. Does it not matter if you have cores to spare? I guess I should read up before I make another stupid statement, thanks all Edit: bought the above book HalloKitty fucked around with this message at 09:15 on Jun 7, 2012 |
# ? Jun 7, 2012 08:41 |
|
the spyder posted:My sales rep just tried to sell me vSphere standard edition... For my two, 192gb ram, dual 8 core hosts... He claimed it supported HA and vMotion, and that I only needed one license... HalloKitty posted:Does it not matter if you have cores to spare? DrOgdenWernstrom posted:http://www.vmware.com/products/datacenter-virtualization/vsphere/small-business/compare-kits.html evil_bunnY fucked around with this message at 10:15 on Jun 7, 2012 |
# ? Jun 7, 2012 10:10 |
|
Also remember that this is a vRAM entitlement, not a physical memory entitlement for the host. That means you have some room to keep free RAM for HA failover and whatever else you need it for.
|
# ? Jun 7, 2012 15:16 |
|
Misogynist posted:Also remember that this is a vRAM entitlement, not a physical memory entitlement for the host. That means you have some room to keep free RAM for HA failover and whatever else you need it for.
|
# ? Jun 7, 2012 15:18 |
|
HalloKitty posted:Does it not matter if you have cores to spare? I guess I should read up before I make another stupid statement, thanks all When you read that you should allocate the minimum number of cores necessary, it's not just a suggestion. We size based on MHz. If a single core VM is running 50% CPU or higher for extended amounts of time, and the application is multithreaded, we will add a core. Otherwise it is just 1 core, no matter what. We regularly tell application vendors to gently caress off on their requirements.
|
# ? Jun 7, 2012 23:06 |
|
adorai posted:Define cores to spare? If you have less vCPUs allocated than you have cores, then you will see no performance penalty. If, like the rest of the world, you are stacked up 5+ vCPUs to cores, you will see a penalty. A VM cannot run until it has all of the cores allocated to it available. So if you have a 2 vCPU VM that needs to run some work, and only 1 core is available, it will wait for the second core to become available. It will hold the idle core during this time. VMware's relaxed coscheduling can fudge this a bit, but for every clock cycle that one core runs, a second core must run as well. Makes sense to me, and it's obviously something I hadn't thought closely enough about. We have a Citrix VDI box that needs reconfiguring a bit based on this (that said, I didn't configure it, but I'll definitely rectify that!) By spare I mean simply that your vCPUs aren't exceeding physically available cores. I realise this probably isn't the most common of scenarios, but I do have a number of machines with 8-12 cores and a load that I can spread pretty evenly over them.
|
# ? Jun 7, 2012 23:12 |
|
The only thing I will give 2 vCPUs on without doing metrics first is Terminal Servers / XenApp servers.
|
# ? Jun 7, 2012 23:14 |
|
Speaking of allocating vCPUs, can anyone speak to the best way to do virtual sockets vs virtual cores in vSphere 5? My assumption has always been 1 vCPU, 2 vCPU, 4 vCPU would be 1 socket 1 core, 2 socket 1 core, 2 socket 2 cores respectively, but it may not actually make a difference. I was at a VMUG today and someone asked that question and the presenter didn't know. My only thought is that two sockets would allow the VM to straddle physical CPUs in the case of NUMA?
|
# ? Jun 7, 2012 23:25 |
|
I benched importing/exporting ~ 10GB MySQL databases - 2 cores vs 4 cores and the 2 core won every time, by more than I thought. I almost want to say the 1 core VM was a hair quicker as well.
|
# ? Jun 7, 2012 23:30 |
|
Erwin posted:Speaking of allocating vCPUs, can anyone speak to the best way to do virtual sockets vs virtual cores in vSphere 5? My assumption has always been 1 vCPU, 2 vCPU, 4 vCPU would be 1 socket 1 core, 2 socket 1 core, 2 socket 2 cores respectively, but it may not actually make a difference. I was at a VMUG today and someone asked that question and the presenter didn't know. There's no difference between sockets or cores, it's just a way to tinker with licensing and compatibility for the guest.
|
# ? Jun 7, 2012 23:32 |
|
I have precisely one VM with >1 vCPU in my environment, and it's a Dynamics AX AOS that will absolutely hammer the poo poo out of a single vCPU. Everything else gets 1 vCPU.
|
# ? Jun 8, 2012 00:18 |
|
adorai posted:Define cores to spare? If you have less vCPUs allocated than you have cores, then you will see no performance penalty. If, like the rest of the world, you are stacked up 5+ vCPUs to cores, you will see a penalty. A VM cannot run until it has all of the cores allocated to it available. So if you have a 2 vCPU VM that needs to run some work, and only 1 core is available, it will wait for the second core to become available. It will hold the idle core during this time. VMware's relaxed coscheduling can fudge this a bit, but for every clock cycle that one core runs, a second core must run as well. Relaxed co-scheduling only "fudges it a bit" if you're underutilizing most of your cores by, say, hammering a multi-CPU VM with a heavy-computing single-threaded workload. This causes stop-start-stop-start effects like shoe-shining an LTO tape. Most of the time, it is a really, really significant SMP performance boost over hypervisors that don't implement relaxed co-scheduling, precisely because it doesn't need to do what you're describing as long as the CPUs are reasonably equally loaded. Vulture Culture fucked around with this message at 04:26 on Jun 8, 2012 |
# ? Jun 8, 2012 04:24 |
|
luminalflux posted:Is there a way to extract performance metrics from vCenter to another graphing tool? I'd like to take the graphs I can see in vSphere Client under the "Performance" tab and get them into something like Graphite or Munin. I looked a bit at SNMP, but vCenter only has traps, no performance counters at all. I know I'm a little late here, but PowerCLI has the get-stat command, if you do a little googling there's some great posts by LucD on using it. You can also use get-view on virtualmachine and gather the VM summary, which will always have quickstats and is very fast to return. There are also pretty simple techniques to gather the realtime stats as well. I wrote a bespoke script with it to gather VM cpu/mem stats from 17 vCenters and dig out what needed rightsizing, so it'll probably provide you what you need.
|
# ? Jun 10, 2012 18:39 |
|
Ok, so fun Citrix fact I discovered this week, related to my earlier posts: If you operate a disjointed namespace environment (that is, the suffix of your computers' FDQN doesn't always match your domain) the desktop agent will try to check in with DESKTOPDIRECTOR.my.domain.com instead of DESKTOPDIRECTOR.disjoint.domain.com, no matter what you set in configuration. If you can't get your connection broker to be in ADDNS, alter the hosts file on your master VM to create a static route. edit: disjointed namespace, not splitscope. sorry, long week. Guesticles fucked around with this message at 20:36 on Jun 10, 2012 |
# ? Jun 10, 2012 20:03 |
|
Mausi posted:I know I'm a little late here, but PowerCLI has the get-stat command, if you do a little googling there's some great posts by LucD on using it. I later found a post on how to get stats from PowerCLI into Graphite, which is exactly what I was looking for. I would have preferred VMware exposing that over SNMP or the Ruby interface having support for get-stat, instead of having to use yet another system to get stats in, but this works.
|
# ? Jun 11, 2012 00:05 |
|
luminalflux posted:I later found a post on how to get stats from PowerCLI into Graphite, which is exactly what I was looking for. I would have preferred VMware exposing that over SNMP or the Ruby interface having support for get-stat, instead of having to use yet another system to get stats in, but this works. Most of the effort with PowerCli isn't getting it working, it's making it run in less than 4 hrs and 4GB of ram.
|
# ? Jun 11, 2012 20:33 |
|
So this continuously pisses me off. Everytime I try and upload a file to my datastores on my ESX 4.1 servers, the vSphere client crashes. OK, fine, I'll SCP the file over. Oh wait, everything is owned by root:root with 755 perms. My guess is that it's crashing on a permission denied error, so how do I fix that? Do I modify the datastore somehow? Do I have to give myself additional permissions? Should the changes I make be on the server via ssh (I doubt it) or from the vSphere client?
|
# ? Jun 11, 2012 22:17 |
|
Mausi posted:I'm glad you found a way to get it working, but that dude's script is massively inefficient. He shouldn't be running a get-stat for each host individually unless that's the only way to get realtime stats; much faster to run it for all hosts by feeding get-stat a list of entities, then running a group-object by entityId afterwards to process on. Looks like I can feed a collection, but as for extraction from the collection, no idea. I'm a total powershell newbie and I can't even program in it effeciently since for some reason typing $ and | through the RDP session from my mac gets mangled. I found LucDs posts though and probably should get a windows VM installed so I can play with this stuff.
|
# ? Jun 12, 2012 00:12 |
|
Put the vSphere client on your bootcamp partition and put the vm on its own desktop, and pronto! Then ctrl+arrows to switch desktops.
|
# ? Jun 12, 2012 01:31 |
|
We're running hyper-v server core on a host with two active guests, both running server 2008r2 standard. One of them, the file/dns/dhcp server, locks up once a month or so. No clients can connect to it, rdp doesn't work but the crappy thing is that it also does not log anything in the eventlogs neither on guest or host (which is weird). We end up killing it through the hyper-v manager. (the second guest is running but not used by anyone) What/Where should i start to look? a systeminfo > out.txt for the host is here http://pastebin.com/kxcuEUxt and for the guest http://pastebin.com/EyH7KE89 What free alternatives are there for monitoring this thing next time it locks up? ** edit I'm an idiot, i give troubleshooting information like i'm one of my users. The server is an IBM x3650 m3 with a lsi-made IBM ServeRAID M5015 raid controller, two 136GB 6gb/s SAS drives in raid 1 for the hosts's system, two in raid 1 for the guests .vhds and two regular sata 1tb drives which the files the fileserver-guest is serving are placed. The megaraid tool says that everything's running fine. Apparently the fileserver responded to ping this morning when it locked up, but nothing more than that. All integration-services are active. ** another edit Looking at scheduled tasks on the client, there's one that ran at 01:30 this morning and the one at 05:00 didn't. At 02:00 Symantec BackupExec is scheduled to do a straight file-backup, INSIDE the guest with no host/hyper-v knowledge at all. Can that be it? There's a start at 0200 and a end at 0200,so something in it was aware of something at that time. It's run fine for the last month.. underlig fucked around with this message at 13:35 on Jun 12, 2012 |
# ? Jun 12, 2012 08:59 |
|
Amazon is amazing in every way, the recommendations system is uncanny:
|
# ? Jun 12, 2012 22:51 |
|
HalloKitty posted:
I was thinking about getting this book, but one if the reviews keeps bringing up that it is quite far out of date. Would this be a good recommended read still, or is there something better out there? (Already have "Mastering vSphere 5" and "HA and DRS technical deepdive")
|
# ? Jun 12, 2012 23:35 |
|
evil_bunnY posted:Put the vSphere client on your bootcamp partition and put the vm on its own desktop, and pronto! I actually caved and did this. I feel sooo dirty. (Fusion has seamless windowing as well, which is still pretty cool even though it's been around for a long time)
|
# ? Jun 13, 2012 00:21 |
|
I didn't see a clearly defined answer here but maybe I missed it. My understanding of the VMware Essentials Plus package was a 192GB VRAM limit across 3 hosts, 6 processors, and you may only allocate a maximum of 32GB of VRAM *per instance*. Yes?
|
# ? Jun 13, 2012 00:39 |
|
I believe its 32GB per socket, so if you have a dual socket server like a R620 or whatever you can have up to 64 GB. Technically you can have more I think since it takes the total amount across all licensed hosts and it cant be more than 192GB across 6 sockets but maybe that is just for standard or enterprise and not essentials.
|
# ? Jun 13, 2012 00:52 |
|
Edit: Reading the lingo on the VMware Website it says "VRAM Entitlement is the amount of virtual memory configured to a virtual machine." So my original understanding appears to be correct. It is 32GB per running virtual instance, 192GB VRAM total allotment. Not per socket. Digital_Jesus fucked around with this message at 12:32 on Jun 13, 2012 |
# ? Jun 13, 2012 00:55 |
|
Moey posted:I was thinking about getting this book, but one if the reviews keeps bringing up that it is quite far out of date. It's mostly about concepts and not specifics. So there's a chapter on how to size your VMs, a chapter on storage concepts, network concepts, etc etc. I've got it already, and it was a pretty quick read. If nothing else, it can be good to put in front of a boss to show them how stupid their idea is.
|
# ? Jun 13, 2012 01:59 |
|
|
# ? May 15, 2024 04:14 |
|
The very concept of virtual RAM per physical socket doesn't even make sense.
|
# ? Jun 13, 2012 04:19 |