|
I should probably swing by the expo and see what's up.
|
#
?
Aug 28, 2012 05:31
|
|
|
|
| # ? Apr 25, 2024 23:02 |
|
|
DevNull posted:Also, I am going to be at VMworld tomorrow. I'll be wearing a blue button up with short sleeves and pink/white stripes. I have a giant octopus tattoo on my right forearm, so I am pretty easy to pick out of a crowd. Say hi, and I will give you all the good VMware gossip. Let me know where you will be and i'll say hi. I'm skipping the morning General Session tomorrow, don't care about workforce/mobile/BYOD/DVI, I'm strictly server side. I've got some general sessions tomorrow afternoon (various Best Practices mainly) then will be at the hall crawl.
|
|
#
?
Aug 28, 2012 07:44
|
|
|
Mierdaan posted:Anyone have the vSphere client working well in Windows8 yet? The install fails saying it requires XP SP3 or higher. There's some forums posts saying you can get it somewhat working with XP compatibility mode, but some features like connecting to consoles and mounting local DVD drives are broken. I just ran the "troubleshoot compatibility" and installed it with the default XP SP3 mode. Haven't had an issue at all, haven't tried to map a DVD though. Console works fine. vSphere 4.1 this is.
|
|
#
?
Aug 28, 2012 12:07
|
|
|
Mully Clown posted:I just ran the "troubleshoot compatibility" and installed it with the default XP SP3 mode. Haven't had an issue at all, haven't tried to map a DVD though. Console works fine. vSphere 4.1 this is. Yeah I get a "The VMRC console has disconnected...attempting to reconnect." error when I try to use a console. vCenter 5.0 U1.
|
|
#
?
Aug 28, 2012 14:34
|
|
|
Current project, replace an ESX3 machine with ESXi5 
|
|
#
?
Aug 28, 2012 14:56
|
|
|
FISHMANPET posted:Current project, replace an ESX3 machine with ESXi5 Was this one of those stories where a server gets bricked inside of some old closet and is just recently discovered? For content: Are those licensing changes they announced effective immediately on ESXi 5.x, or does it come into play once 5.1 is released? If it is in effect now, I am tempted to update my hosts from 4.1 to 5.0 sooner than later. Moey fucked around with this message at 15:28 on Aug 28, 2012 |
|
#
?
Aug 28, 2012 15:22
|
|
|
Misogynist posted:If it does end up being great then I'm a little pissy that we made a huge initial investment in PHD Virtual this year, but something that's able to back up 1.2 TB of VMs using less than 8,000,000 inodes on its NFS share would sure be handy from an administration perspective. I'm sure they'll be playing feature-catchup for awhile yet, though. Heard a sad rumor that PHD Virtual was running low on cash.
|
|
#
?
Aug 28, 2012 15:27
|
|
|
Moey posted:Was this one of those stories where a server gets bricked inside of some old closet and is just recently discovered? No, it's a single shot that was set up in a research lab a million years and nobody ever looked at it again. I'm hoping to buy a second server and setup vCenter for them, since during experiments there has to be 100% uptime.
|
|
#
?
Aug 28, 2012 17:07
|
|
|
Got a question for people that know far more about this than I do. I have a couple windows server 2008 R2 x64 VMs running on a Cisco UCS blade (which is plenty fast, huge amount of memory etc). Storage is provided by a NetApp filer (don't have the exact spec here). They are NFS datastores and we're using 10gb ethernet all the way from the server through to the filer. The disk array has something like 60 10k (?) SATA drives in it and is used fairly lightly from what I understand. I am running iometer on the vm (50% read with 4kb block size, but I tried a bunch of other tests also) and seeing approx 60 Mbps max throughput with 8 concurrent worker threads. I was expecting to see higher than this honestly - it seems very low. Is there any suggestions about what I ask our server/network/storage guys to look at for this?
|
|
#
?
Aug 28, 2012 17:21
|
|
|
At 4KB blocks you're never going to see huge amounts of data. Figure out what drives you actually have, and whether the volumes you're writing too are on an aggregate that spans all of them. E: 60MB/s is 1500 iops at 4k, which is about what a shelf of 7,2k drives should be capable of. Until you get the filer config there's not much to tell. evil_bunnY fucked around with this message at 18:50 on Aug 28, 2012 |
|
#
?
Aug 28, 2012 17:27
|
|
|
evil_bunnY posted:At 4KB blocks you're never going to see huge amounts of data. Im quite embarrassed to say this but I misconfigured IOMeter (I somehow only configured 1 worker thread not all of them). Now I'm seeing considerably better results. Using the default 32kb 50% read test shows average around 400mb/s. Same config except 4kb is sitting around 200.
|
|
#
?
Aug 28, 2012 20:38
|
|
|
Measuring throughput instead of operations at low blocksizes is still kind of misleading (yourself).
|
|
#
?
Aug 28, 2012 21:00
|
|
|
Erwin posted:Is this for View? We recently did an initial test of View, and after 30 seconds of dragging a window around the screen, we threw up a little in our mouths and put it on the back burner. Confirmed this. 1 Gb link is the min best practice, recommend multi 10 Gb links. Only removes the need for shared storage. Same vCenter is required. It is in the future path to do cross vCenter, cross VDC, and long distance vMo but not now.
|
|
#
?
Aug 28, 2012 22:06
|
|
v
|
evil_bunnY posted:Measuring throughput instead of operations at low blocksizes is still kind of misleading (yourself). Yep. I also just realised that IOmeter is showing ~60,000 iops for that test. I think that is probably.. slightly unrealistic... unless there is all kinds of caching going on there somewhere. Spoke with our storage architect and we're going to take a deeper look tomorrow. As someone with a dev/apps support background a lot of the infrastructure stuff is really new to me, so this is a huge educational experience for me.
|
|
#
?
Aug 28, 2012 22:12
|
|
|
Can I join my ESX3 server into a cluster with an ESXi5 and vMotion the guests from the ESX3 server to the ESXi5 server? Once they migrate off the ESX3 server will be retired, but if I can do the migration seamlessly I'll be a superstar hero.
|
|
#
?
Aug 28, 2012 22:30
|
|
|
Yes, with some caveats.
|
|
#
?
Aug 28, 2012 22:45
|
|
|
It's how most migrations are done, but it's also a situation where skimping on the due diligence will gently caress you over in new and interesting ways.
|
|
#
?
Aug 28, 2012 23:33
|
|
|
So ESX3 supports virtual machine hardware version 4, as does ESXi5. And ESX3 supports VMFS3, which ESXi5 does as well, so I should be fine on that front (with the caveat that I should upgrade the hardware to version 8 when I can, and svMotion everything to a VMFS5 volume). Is there anything else I need to be concerned about?
|
|
#
?
Aug 29, 2012 01:29
|
|
|
FISHMANPET posted:with the caveat that I should upgrade the hardware to version 8 when I can
|
|
#
?
Aug 29, 2012 02:57
|
|
|
FISHMANPET posted:Current project, replace an ESX3 machine with ESXi5
|
|
#
?
Aug 29, 2012 22:19
|
|
|
And here I thought I was behind the times running 4.0!
|
|
#
?
Aug 29, 2012 23:25
|
|
|
I'm so far behind that most of my ESX hosts are actually Hyper-V.
|
|
#
?
Aug 30, 2012 01:17
|
|
|
I left a 3.5 box running at my old employer 
|
|
#
?
Aug 30, 2012 02:07
|
|
|
Mausi posted:I'm still retiring ESX3.x hosts, you're not alone good sir. Our VMware hosts are at ESX 3.5. We are migrating them over to our XenServer farm since we already use that for our Citrix infrastructure.
|
|
#
?
Aug 30, 2012 04:52
|
|
|
What exactly are CPU requirements for FT? I'm looking at running an FT VM on a host with a 55xx chip and another host with an E5-24xx chip. Does that use EVC to properly mask the correct CPU bits so it all runs at the 55xx level? I can't find anything definitive in the recommendations, other than that the CPUs need to be compatible, but there's no explanation of what that means.
|
|
#
?
Aug 30, 2012 07:47
|
|
|
FISHMANPET posted:What exactly are CPU requirements for FT? I'm looking at running an FT VM on a host with a 55xx chip and another host with an E5-24xx chip. Does that use EVC to properly mask the correct CPU bits so it all runs at the 55xx level? I can't find anything definitive in the recommendations, other than that the CPUs need to be compatible, but there's no explanation of what that means. I swear, VMware's documentation is actually really good!
|
|
#
?
Aug 30, 2012 14:34
|
|
|
Mierdaan posted:I swear, VMware's documentation is actually really good! That hasn't been updated since October, so it doesn't include any modern CPUs (like the E5-24xx I'll be using), and also isn't very clear. Do all the CPUs have to be under the same bold heading, or do all the CPUs have to be on that list?
|
|
#
?
Aug 30, 2012 14:37
|
|
|
Same bold heading. Here's an updated list that includes Sandy Bridge procs, but still not the E5-24xx specifically. Download the Site Survey tool if you want to be sure.
|
|
#
?
Aug 30, 2012 15:00
|
|
|
That's about EVC for vMotion, nothing about FT. And C'mon VMWare, I'd like to know if this will work before I spend $50k on it, not after. That Site Survey tool is pretty stupid in that context.
|
|
#
?
Aug 30, 2012 17:09
|
|
|
That's about EVC because FT depends on HA (can't power on an FT VM without HA enabled) and on EVC for DRS (otherwise DRS is disabled for FT VMs).
|
|
#
?
Aug 30, 2012 17:13
|
|
|
Mierdaan posted:That's about EVC because FT depends on HA (can't power on an FT VM without HA enabled) and on EVC for DRS (otherwise DRS is disabled for FT VMs). 11 acronyms and 18 words. Welcome to VMware.
|
|
#
?
Aug 30, 2012 17:17
|
|
|
Holy poo poo that sentence.
|
|
#
?
Aug 30, 2012 17:18
|
|
|
DevNull posted:11 acronyms and 18 words. Welcome to VMware. Yes but none of those had a lower case 'v' in front of them, progress!
|
|
#
?
Aug 30, 2012 17:19
|
|
|
DevNull posted:11 acronyms and 18 words. Welcome to VMware. Who would have thought virtualizing digital computing devices would involve so much abstraction?
|
|
#
?
Aug 30, 2012 17:26
|
|
|
So the reason I ask, is a coworker who is a programmer but knows just enough to be dangerous did some research on my plans to use FT between the R710 and the R520, and found a vSphere 4 FT FAQ that included the following:quote:ESX host processors must be VMware FT capable and belong to the same processor model family. VMware So I understand what EVC and what it does, but EVC has been around for a long time yet the FAQ makes no mention of it and basically says the CPUs have to be in the same family. So either that's still true, in which case it would be nice to find a vSphere 5 FT FAQ that mentions that, or it's not true and EVC can keep CPU properly masked, in which case it would be nice to find a document that says "Hey, we removed this limitation!"
|
|
#
?
Aug 30, 2012 21:13
|
|
|
FISHMANPET posted:So the reason I ask, is a coworker who is a programmer but knows just enough to be dangerous did some research on my plans to use FT between the R710 and the R520, and found a vSphere 4 FT FAQ that included the following: Just so we're all clear here, you know that FT is not usually a good idea? Unless you happen to have a 1-core VM that for some reason needs an insane amount of uptime, the overhead and configuration you'll have to deal with for FT is usually wasted.
|
|
#
?
Aug 30, 2012 21:51
|
|
|
madsushi posted:Just so we're all clear here, you know that FT is not usually a good idea? Unless you happen to have a 1-core VM that for some reason needs an insane amount of uptime, the overhead and configuration you'll have to deal with for FT is usually wasted. Yeah, I'm well aware. This is for a research lab that does experiments that cost hundreds of thousands of dollars to run. If a server collecting data goes down during an experiment then you've lost that data and have to rerun the experiment. I'm hoping what they need protected is 1 core, and if not I'll figure out how to reduce it to 1 core. And if it ends up being that we need two identical servers then we can buy another server, I just need to know if that's necessary or not.
|
|
#
?
Aug 30, 2012 22:17
|
|
|
vSMP FT shouldn't be too far away. They're demo'ing it in top secret labs now.
|
|
#
?
Aug 30, 2012 22:19
|
|
|
three posted:vSMP FT shouldn't be too far away. They're demo'ing it in top secret labs now.
|
|
#
?
Aug 30, 2012 22:23
|
|
|
|
| # ? Apr 25, 2024 23:02 |
|
|
In unrelated news, curse you VMware for moving Storage vMotion into Standard, it meant I had to defend myself against a claim of being lazy from my director for wanting DRS in Enterprise (though I think it may have been a devil's advocate thing from the director, but the managers didn't get that memo, so gently caress that discussion). E: And unrelated, to possibly play devil's advocate with myself, any thoughts on getting production vs basic SNS vs only getting the first required year and flying by the seat of your pants? E2: And to clarify, I'm actually working in two seperate VMware deployments currently, so if some of my questions seem like they don't all mesh together, it's because they don't. Like saving money on support for one cheap department vs dumping money into hardware for another rich department. FISHMANPET fucked around with this message at 22:34 on Aug 30, 2012 |
|
#
?
Aug 30, 2012 22:30
|
|