Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug
Yeah pretty much agree with Three, most of the UCS stuff is things you can already, or free, do if you know how to utilize your feature set given.

adorai posted:

List price might be more expensive, they are reasonably competitive with normal discounts.

Yeah the Gen3's are reasonably priced, smartnet renewals can be a pain sometimes, with promotions that never seem to end, I do like if you are just unhappy with the server TAC will basically send you a full new one.

I do wish they would use some AMD procs in their UCS line up however

Dilbert As FUCK fucked around with this message at 03:29 on Apr 4, 2013

Adbot
ADBOT LOVES YOU

three
Aug 9, 2007

i fantasize about ndamukong suh licking my doodoo hole

Corvettefisher posted:

I do wish they would use some AMD procs in their UCS line up however

Is that for technical or philosophical reasons?

1000101
May 14, 2003

BIRTHDAY BIRTHDAY BIRTHDAY BIRTHDAY BIRTHDAY BIRTHDAY FRUITCAKE!

three posted:

The hardware detaching is cool, but how often does anyone need to use it? It'd be easier to just use auto deploy for ESXi hosts, and only bad people run physical workloads nowadays. (Plus who boots Windows from SAN? Although, technically you could just swap the drives.)


I use it pretty frequently. I also like the idea of policy based management and being able to manage 200 blades from a single interface without having to buy a manager of managers.

It's also pretty rad when you can sort your blades into different organizations and then apply permissions/policies. So if you have 100 blades you can give say 20 to dev, 30 to QA and the rest to production. Then the Dev/QA engineers could have total control over their blades without being able to mess around with production stuff.

Taking that one step further it opens up the option for production or QA or Dev to borrow resources back and forth without having to deal with re-installation/deployment if you don't want to.

It's also nice to pop in a new chassis then go click on your service profile template and tell it you want 8 more ESX servers configured in exactly the way you need. As an example our ESXi servers tend to have 6 NICs configured (lab machines with a pair of NICs on a standard switch, a pair on a dvSwitch and a pair on the Nexus 1000v) and our hyper-v systems only have 4.

I find UCSM is a lot friendlier when you have multiple administrators poking around doing stuff. If two people try to open a KVM console for the same blade it'll offer a chance to share it. Last time I dealt with Dell, HP and IBM they usually just said "sorry this is locked! try again later!" which was really frustrating when I was the one using it but my browser crashed from trying to use virtual media. This may have improved in recent versions. I haven't had a chance to touch that Dell M1000 in the lab.

quote:

It would be better to use NIOC.

I'd prefer real QoS to NIOC. Mostly by virtue of QoS being honored outside of the dvSwitch.

Vulture Culture
Jul 14, 2003

I was never enjoying it. I only eat it for the nutrients.

three posted:

Is that for technical or philosophical reasons?
One obvious reason: vMotion can only happen between processors of the same manufacturer, so if someone's running an AMD shop already, you can't run a (robust) mixed cluster with UCS in the middle.

DagPenge
Jun 4, 2011

Looks like our civilians are fine, thank god for the capitalist spirit!
Was thinking about testing Microsofts Hyper-v solution, since it is way cheaper than VMware. Anyone know of anything I should be aware of? Also can you run the SQL you need for System center VMM virtual on the same cluster it is supporting?

Syano
Jul 13, 2005

DagPenge posted:

Was thinking about testing Microsofts Hyper-v solution, since it is way cheaper than VMware. Anyone know of anything I should be aware of? Also can you run the SQL you need for System center VMM virtual on the same cluster it is supporting?

Whats your deployment scenario? Hyper-v isn't always 'way cheaper'. I mean it can be but it depends. And yes you can run your scvmm instance as a VM.

DagPenge
Jun 4, 2011

Looks like our civilians are fine, thank god for the capitalist spirit!
For a small setup thats supposted to have AD, Exchange, SQL and a few hosted desktop (terminal services or whatever its called now) for about 10-20 users each. Hyper-v seems to be the easiest way to get going, since it's included with the Microsoft licenses I'll need anyway.

Is this easy to visualize or should I make up a drawing to better explain?

Kachunkachunk
Jun 6, 2011

evil_bunnY posted:

Is it ok to admit I love vDS'es mostly because I'm lazy?
Netflow is really cool as well!

Mierdaan
Sep 14, 2004

Pillbug
Joy: build 5.0.0.913577 of the vSphere client fixes the console nonsense on Windows8. No more using Workstation 9 for console access.

DevNull
Apr 4, 2007

And sometimes is seen a strange spot in the sky
A human being that was given to fly

Mierdaan posted:

Joy: build 5.0.0.913577 of the vSphere client fixes the console nonsense on Windows8. No more using Workstation 9 for console access.

I actually use WS most of the time. The only thing I need but can't do is add VMs to my inventory from the datastore. I have been bugging the product manager about it. Hopefully, it will be next release.

Kachunkachunk
Jun 6, 2011

Mierdaan posted:

Joy: build 5.0.0.913577 of the vSphere client fixes the console nonsense on Windows8. No more using Workstation 9 for console access.
What kind of console nonsense? I'm using Windows 8 as well but I find the console remains blank when starting or restarting VMs until I close and re-open the console a second time.

Mierdaan
Sep 14, 2004

Pillbug
"The VMRC console has disconnected...attempting to reconnect" for all consoles all the time, was the behavior I and many others got.

mAlfunkti0n
May 19, 2004
Fallen Rib
Does anyone here have any background scripting esxcli commands on 4.1 hosts?

We are doing a massive storage migration and each LUN I remove from the host has to me masked (see below) :

esxcli corestorage claimrule add --rule 110 –t location –A vmhba3 –L 156 –P MASK_PATH
esxcli corestorage claimrule add --rule 111 –t location –A vmhba4 –L 156 –P MASK_PATH

I have to repeat this for each LUN being removed. Is there an easy way to script this process?

Edit : BTW, 5.0 makes this process much easier and is all done via the GUI with a simple right click. 4.x is a miserable beast, especially when you have to migrate about 100 LUNs. Shoot me now.

Edit 2: We were able to script it (bash script) easily, I would love to see if we can do this from power cli.

mAlfunkti0n fucked around with this message at 19:47 on Apr 4, 2013

Goon Matchmaker
Oct 23, 2003

I play too much EVE-Online
Handy script I wrote earlier:

code:
#!/bin/bash
for h in <hosts here>; 
do
ssh root@$h 'for i in `ls /vmfs/devices/disks/ | grep naa.600` ; do esxcli storage nmp device set --device $i --psp VMW_PSP_RR; done'
ssh root@$h 'for i in `ls /vmfs/devices/disks/ | grep naa.600` ; do esxcli storage nmp psp roundrobin deviceconfig set -d $i --iops 1 --type iops;done'
done
Goes through each LUN and sets the native multipathing to use round robin then adjusts each lun to shoot one iops down a path before switching to the other one. With this setup we got a nice increase in performance over the default of 1000 iops before switching paths.

Frozen Peach
Aug 25, 2004

Heroes Never Die

When it comes to virtualizing things like exchange that use all the ram you give them, how do I know if I should allocate more memory to the VM?

Right now our new server is floating at 8.3-8.4 GB used out of the 10 we gave it. Everything seems to be running fine. It's a small server with 75 users and 25 GB of mailboxes. I'd like to say that we are fine since we're not at 100% memory on the VM. I'm just not sure how to tell definitively.

Maneki Neko
Oct 27, 2000

Frozen-Solid posted:

When it comes to virtualizing things like exchange that use all the ram you give them, how do I know if I should allocate more memory to the VM?

Right now our new server is floating at 8.3-8.4 GB used out of the 10 we gave it. Everything seems to be running fine. It's a small server with 75 users and 25 GB of mailboxes. I'd like to say that we are fine since we're not at 100% memory on the VM. I'm just not sure how to tell definitively.

This may depend on the application honestly, there may be different sets of perfmon counters, etc. to check to determine if you're experiencing memory pressure or not.

bull3964
Nov 18, 2000

DO YOU HEAR THAT? THAT'S THE SOUND OF ME PATTING MYSELF ON THE BACK.


Frozen-Solid posted:

When it comes to virtualizing things like exchange that use all the ram you give them, how do I know if I should allocate more memory to the VM?

Right now our new server is floating at 8.3-8.4 GB used out of the 10 we gave it. Everything seems to be running fine. It's a small server with 75 users and 25 GB of mailboxes. I'd like to say that we are fine since we're not at 100% memory on the VM. I'm just not sure how to tell definitively.


That's when you start looking at the application performance counters and see if they indicate RAM pressure. Specifically, the Memory Paging Counters are a good place to start.

http://technet.microsoft.com/en-us/library/ff367896(v=exchg.141).aspx

Vulture Culture
Jul 14, 2003

I was never enjoying it. I only eat it for the nutrients.

bull3964 posted:

That's when you start looking at the application performance counters and see if they indicate RAM pressure. Specifically, the Memory Paging Counters are a good place to start.

http://technet.microsoft.com/en-us/library/ff367896(v=exchg.141).aspx
This will tell you gently caress-all when you're running in a virtualized environment with a balloon driver, though. By increasing your RAM you might actually be increasing how much you swap.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug

Frozen-Solid posted:

When it comes to virtualizing things like exchange that use all the ram you give them, how do I know if I should allocate more memory to the VM?

Right now our new server is floating at 8.3-8.4 GB used out of the 10 we gave it. Everything seems to be running fine. It's a small server with 75 users and 25 GB of mailboxes. I'd like to say that we are fine since we're not at 100% memory on the VM. I'm just not sure how to tell definitively.

Exchange likes to cache, depending on what you are making it do it can take a lot. If you are on standard or higher you can always hot add ram, so you can start low and gradually add more if ram is being the bottleneck.. Generally I'll run VM's at 65-80% usage or vRam.

Incase you are wondering here is the white paper for Exchange 2010
http://www.vmware.com/files/pdf/exchange-2010-on-vmware-best-practices-guide.pdf

Frozen Peach
Aug 25, 2004

Heroes Never Die

The good news is we're not even CLOSE to our full RAM utilization. Even if every server that was allocated memory took 100% it was allocated, we'd be barely over 60% total utilization on that VM host. We're sitting at 17 GB total allocated out of 32 GB on that blade. I can afford to allocate more, but I only want to do it if I have to. Right now I can fit several VMs from our other host on this, if the other host goes down spontaneously. We're not at risk of swapping at all.

I'm just at a loss when Exchange will take literally every byte of RAM it can just to cache, but I don't have a clue how to tell if the RAM is bottlenecking us. That's where I start getting out of my league.

Frozen Peach fucked around with this message at 04:18 on Apr 5, 2013

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug
Swapping being reported from ESXi is the LAST thing you want to see, Ballooning and Compression should be warning signs that your host is running into oversubscribed ram issues. Transparent page sharing, is a great function which helps lower ram usage on the host as well.

Ram would be the bottleneck if you are seeing the guest page it's memory into swap, usage runs ~90% during normal operation, the windows OS complains about memory via event logs or warnings, and/or performance is poo poo when you can verify IOPS, CPU, and network are not constrained.

Mierdaan
Sep 14, 2004

Pillbug
How much of the 5.1 upgrade complexity does this help with? I haven't been through it yet.

http://blogs.vmware.com/kb/2013/04/introducing-the-vcenter-certificate-automation-tool-1-0.html

sanchez
Feb 26, 2003
Exchange does start spewing log errors if it gets really low on RAM, I've only seen it on single server 2010 environments with 4gb of ram though (not my idea). Once you get up past 8-10gb it hums nicely, I suspect (un-scientifically) based on your environment size it'll be just fine with 10.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug

Mierdaan posted:

How much of the 5.1 upgrade complexity does this help with? I haven't been through it yet.

http://blogs.vmware.com/kb/2013/04/introducing-the-vcenter-certificate-automation-tool-1-0.html

It's honestly pretty straight forward so long as you don't select simple install. Also put SSO in multi-site even if you are currently one site, there is no easy way to change it after you finish it out.

Replacing certs in 5.1 was a bit more tedious, but that tool looks pretty cool.

madsushi
Apr 19, 2009

Baller.
#essereFerrari

Corvettefisher posted:

It's honestly pretty straight forward so long as you don't select simple install. Also put SSO in multi-site even if you are currently one site, there is no easy way to change it after you finish it out.

Replacing certs in 5.1 was a bit more tedious, but that tool looks pretty cool.

Yeah, as long as you don't choose the wizard (which should be the safest).
And as long as you choose multi-site (without being warned that you can't change later).
And as long as you don't change your RAM/CPU count after installing (or you'll break the Java thumbprint).
And as long as your password doesn't have a ! or \ or other character in it (since it passes the password via the command line).

The 5.1 upgrade process sucks hard unless you've been through it and figured out all the "gotchas". Enterprise-grade software should not be this difficult to install/upgrade. VMware messed up. Your software shouldn't have "traps" set up that people can fall into.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug
Doesn't installing a latest version of Java reset that foot print? I know it does something to combat the poo poo that vCenter became.

OH LET'S JUST WRAP IT IN JAVA WHAT COULD GO WRONG!

Vulture Culture
Jul 14, 2003

I was never enjoying it. I only eat it for the nutrients.
Anyone know what causes this?



None of the VMs on here are backed up through PHD Virtual, so that's not it.

Edit: This happens on basically all our datastores at the same time, across four different SANs.

Vulture Culture fucked around with this message at 20:06 on Apr 5, 2013

Erwin
Feb 17, 2006

I assume that's in vCenter and not a specific host? I had the same problem and support said it's a bug with vCenter and the way it polls usage. It was annoying because it kept crossing an alarm threshold and I kept getting alarms. I don't remember how I resolved it, but it probably involved rebooting vCenter or some stupid poo poo.

three
Aug 9, 2007

i fantasize about ndamukong suh licking my doodoo hole

Misogynist posted:

Anyone know what causes this?



None of the VMs on here are backed up through PHD Virtual, so that's not it.

Edit: This happens on basically all our datastores at the same time, across four different SANs.

Are you doing any replication outside of PHD Virtual that would invoke snapshots? Any corresponding events on VMs that reside there at that time?

Vulture Culture
Jul 14, 2003

I was never enjoying it. I only eat it for the nutrients.

Erwin posted:

I assume that's in vCenter and not a specific host? I had the same problem and support said it's a bug with vCenter and the way it polls usage. It was annoying because it kept crossing an alarm threshold and I kept getting alarms. I don't remember how I resolved it, but it probably involved rebooting vCenter or some stupid poo poo.
This is a definite possibility.


three posted:

Are you doing any replication outside of PHD Virtual that would invoke snapshots? Any corresponding events on VMs that reside there at that time?
Array-based synchronous replication only.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug
I am going to be covering VCP/VCP-DT on my new site, any fine points I should cover?

Cronus
Mar 9, 2003

Hello beautiful.
This...is gonna get gross.
So I am taking the 5.1 ICM class remotely starting next week. Later in July I may be able to get my company to cover the optimize and scale course for VCAP.

I am wondering if there is any major benefit to the higher level cert other than sending your résumé up a notch? Essentially I'm trying to figure out which path to go for -- focus on VCAP or go for the VCP-DT? I am leaning towards the latter for diversity but I was wondering what others thought.

Currently I am one of the senior admins at a MSP and we are very VMware focused. We have View in production but few people use it.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug

Cronus posted:

So I am taking the 5.1 ICM class remotely starting next week. Later in July I may be able to get my company to cover the optimize and scale course for VCAP.

I am wondering if there is any major benefit to the higher level cert other than sending your résumé up a notch? Essentially I'm trying to figure out which path to go for -- focus on VCAP or go for the VCP-DT? I am leaning towards the latter for diversity but I was wondering what others thought.

Currently I am one of the senior admins at a MSP and we are very VMware focused. We have View in production but few people use it.

I have the VCP/VCP-DT, the VCP-DT the outline for the DT was great the exam felt like "NAME AD POLICIES TO OPTIMIZE PCoIP", when this new job starts I am going to pass my VCAP-DCA as I get a nice bonus! There is also a VCAP-DT, which looks fun, and my site should be covering all the objectives. Video editing on the laptop I have is much slower than I thought. The VCAP is a great course to go over, Troubleshoot especially.

If you are going for view stuff these books are great
http://www.amazon.com/VMware-View-Building-Successful-Technology/dp/032182234X/


For VCAP and such
http://www.amazon.com/VMware-vSphere-Design-Forbes-Guthrie/dp/1118407911/
http://www.amazon.com/Administering-vSphere-Planning-Implementing-Troubleshooting/dp/1435456548/
These books really helped me out.

Just remember if you are unsure the answer or if you have it right move on, because your time zips by.

three
Aug 9, 2007

i fantasize about ndamukong suh licking my doodoo hole
It seems inappropriate to take a certification as a given when not passed, and to provide advice on exams you have not taken.

Dilbert As FUCK
Sep 8, 2007

by Cowcaster
Pillbug

three posted:

It seems inappropriate to take a certification as a given when not passed, and to provide advice on exams you have not taken.

I took the VCAP-DCA and missed it by a stupid close number, if I just didn't spend so much time on autodeploy I would have been fine, and took I should have taken it BEFORE a week in Vegas at PEX. Granted I haven't taken the VCAP-DT but the book I gave was in reference to the VCP-DT, which I did pass.

Red Robin Hood
Jun 24, 2008


Buglord
Our office had our main AD server die today. Is it possible to add a domain controller to a domain when the primary has died?

Am I even in the correct thread? :psyduck:

Walked
Apr 14, 2003

Red Robin Hood posted:

Our office had our main AD server die today. Is it possible to add a domain controller to a domain when the primary has died?

Am I even in the correct thread? :psyduck:

Probably not the right thread.
But regardless: Google how to sieze FSMO roles. It's not too hard

http://support.microsoft.com/kb/255504

Red Robin Hood
Jun 24, 2008


Buglord

Walked posted:

Probably not the right thread.
But regardless: Google how to sieze FSMO roles. It's not too hard

http://support.microsoft.com/kb/255504

We don't want to seize the FSMO role as we don't wan to replace the primary just add another secondary. Does that make sense? If you can think of a more appropriate thread please point me in that direction! Thanks!

Docjowles
Apr 9, 2009

Red Robin Hood posted:

We don't want to seize the FSMO role as we don't wan to replace the primary just add another secondary. Does that make sense? If you can think of a more appropriate thread please point me in that direction! Thanks!

General Windows enterprise IT chat
Group policy/AD thread

Adbot
ADBOT LOVES YOU

evil_bunnY
Apr 2, 2003

Red Robin Hood posted:

We don't want to seize the FSMO role as we don't wan to replace the primary just add another secondary. Does that make sense? If you can think of a more appropriate thread please point me in that direction! Thanks!
This is not how a modern AD works. Go ask in the thread Docjowles linked.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply