Sorry if I caused any confusion. What I said only applied to the exam itself, not to the voucher. If the voucher already expired, you'd unfortunately have to pay for the exam, but you definitely don't have to retake the class.
André
Re: Missing VCP5 Exam
↧
↧
VCOPS for a small env
Hi,
My customer are mostly SMB with 2-3 ESXi host usually less than 100 VM. Do I need as much as 16GB of RAM to run vCOPS ? For some customer vCOPS would be one of the largest workload...
Anyone run vCOPS in a small environment like this ? How much RAM can I expect to use ?
↧
Re: ESXi 5.1 - Disable Flow Control
First of all, as you probably noticed Flow Control on ESXi operates in autonegotiation mode by default, so it will only be active if the physical switch supports and enables it too. Some good information on the topic in general: http://rickardnobel.se/esxi-and-flow-control/
As to why you can't change the parameters, check if there is an updated driver for your NIC. Use ethtool -i vmnicX to see which driver version is currently running.
The ESXi HCL lists a few updated qlcnic async drivers:
The latest one being this qlnic 5.1.157:
https://my.vmware.com/web/vmware/details?downloadGroup=DT-ESXI5X-QLOGIC-QLCNIC-51157&productId=285
qlcnic-esx50-5.1.157-1089431.zip
Release Date: 2013-04-12
Build Number: 1089431
Update the driver and see if you can configure the pause feature then.
↧
Re: Can't get vsphere eval license key
Andre,
I was able to get another eval license for vSphere 5 and ESXi5 from my regional sales rep and was able to add the new eval license to the current platform. My evaluation platform will now expire at the end of August giving me plenty of time to complete the build out of our new permanent environment. And no, it doesn't look like the vSphere 5 and ESXi5 eval licenses are available without going through the sales department.
Thanks again for your help and suggestions. I'm glad I was able to forgo the transfer of the permanent license from the eval platform to the permanent platform.
Much appreciated.
↧
Re: Isolation d'un réseau virtuel
Bonsoir,
Côté réseau un Vlan sera sans doute le plus simple. il suffira d'assurer la translation entre les Vlans depuis un point (un firewallpar exemple sinon ya pleins de trucs qui font ça)
la solution la plus simple ensuite est de jouer sur le vswitch de ton esxi :
1- tu crée 2 groupes de ports sur ton Vswitch avec des noms distincts (genre lan et vlan1) et tu attribue les bon réseaux à tes Vms.
2- tu peu sinon simplement créer un second vswitch.
3-enfin tu map ton adaptateur reseau Physique sur le vswitch (par exemple NIC0 = LAN et NIC1 = Vlan1), et dans le cas ou tu as déclarer un groupe de ports pense à renseigner les Vlans.
Si on suit les préconisations de VMware ton Vmkernel de managment doit être sur Vswitch isolé (et relié à ton Lan .. ), ça c'est à toi de voir si dans l'avenir tu veu faire evoluer ton infra ou pas.
Ce qui fais donc bien au total 3 ports physique ( exemple : NIC0=mgmt, NIC1= LAN, NIC2=Vlan1) pour ton cas.
p-s : surveille bien que tes adaptateurs réseaux (NIC )soient pluggés au bon endroit sur tes switch physiques et que (802.1Q de mémoire soit supporté) le trunk soit correctement configuré (attention encore à l'appelation trunk sur le matos Cisco ^^).
Après ça roule tout seul 
.
Bon Courage et bon weekend ^^
↧
↧
Re: Seperate Cluster for single and dual core VMs
There's no reason the separate the VM's just because of the number of virtual CPUs. I'd definitely go with a single cluster and let the resource scheduler do its job.
André
↧
Re: VM Memory Reservation
johanswa wrote:
PMEM/MB = 2031 vmk
Memory overhead = 1.2GB
So all overhead = 3.9GB.
Now is starting to makes sense. Was cutting it way to close, didnt know about the PMEM/MB vmk value.
Well, we are struggling with SQL performance problems, and SQL prefers a memory reservation, even though ballooning wont take please since there is only one VM on the host?
SQL prefers a memory reservation - Where did you get this idea from? Have you had Microsoft or someone who is an expert in SQL performance take a look at your environment? Keep in mind that just because "Task Manager" says sqlservr.exe is consuming, say 99GB of memory, doesn't mean that it is actually *using* that much memory... it is going to claim at a minimum what the SQL admin set as the minimum memory for the instance, and usually will grab much more, if it is available, for buffer space (call it cache for sake of discussion). When Windows encounters memory pressure, SQL will give some of that memory back.
I agree with jrmunday that you need to consider looking at your I/O. You might have some tweaks either in your storage system, or maybe you have an unoptimized query, or bad application code, or some indexes that need to be created, or something else that causes more I/O than necessary. Managing SQL performance is a complex undertaking that requires expertise from several IS diciplines to handle, at times.
↧
Re: Creating a thinapp from an aplication that is installed from a CD
I tried virtualizing the CD ROM, but what happens is when I click the entry point shortcut on the desktop, the logo appears in the lower right hand corner of the screen, the cursor does its circle thing for a moment, then nothing. No error, just nothing. I had that happen before with trying to virtualize a CD program, but never did figure out what was causing it to not function. The CD files work perfectly when copied to the C drive and run from there, so that is what I was trying to replicate as a ThinApp, rather then bother with virtualizing the CD ROM, but both captures present the same result, just like it's trying to open, then nothing. Thanks.
↧
Re: Metro vMotion Question
As a simple normal vMotion does not change the storage location of the VM, yes. This requires a stretched cluster that is basically completely unaware of the physical separation and distance of the cluster members. This does not require you do have a (replicated) storage array at every location either, you could just build the cluster with a single storage but hosts in different locations as long as they can all access the same storage (and layer 2 network). It's completely transparent in that regard.
The destination host would still access the same LUN on the same array, unless you implement some abstraction or storage-side magic in between.
The first, simple case would be a uniform stretched cluster while another case that keeps the storage access local is a non-uniform stretched cluster.
Read more about this topic here:
http://www.yellow-bricks.com/2011/10/05/vsphere-5-0-ha-and-metro-stretched-cluster-solutions/
↧
↧
Re: Seperate Cluster for single and dual core VMs
Would this mixture increase my ready times and adversely affect performance? If all vms on the hosts have 2 vcpus then as one vm finishes then 2 physical cores will always be available... if you toss single vcpu request into the mix then there may be added wait/ready time while waiting for an opening to process a dual core request. Right?
↧
Oracle RAC VM
Hello,
I am trying to build a VM for Oracle RAC cluster. When adding the disk I suppose to see the option "Support clustering features such as Fault Tolerance".
Do not see it. Missing something.
Appreciate any help.
Thank you
↧
Virtualized MSSQL Servers in mirror failing over during vmotion
Hello we've just finished virtualizing our SQL environment. We use SQL mirroring as a fail over mechanism.
Our DBAs are telling me that when a vmotion occurs it causes the SQL mirror to fail over.
Has anyone seen this issue before, is there some sort of setting that I can tweak in case it's due to an IO timeout?
This is a SQL 2008 R2 environment (I think the latest SP)
Upon starting at this new company I've just finished the migration from ESX 4.1 to ESXi 5.1 Update 1a, all Enterprise Plus licensing.
We have rules in place to separate some of the SQL servers in the mirror, so the VMs themselves don't vmotion too much, usually only when we are doing maintenance.
Any help or guidance would be helpful, thanks!
↧
Re: Seperate Cluster for single and dual core VMs
Take a look at Contention on lightly utilized hosts and the links within this blog post, which explain how co-scheduling was improved in version 4.1. If you don't run on e.g. a host with a single dual-core CPU you don't have to really worry about ready times with the VM specifications you mentioned. I'm usually running 1 to 4 vCPU VMs on different environments/hosts without issues.
André
↧
↧
Re: Error While connecting to the vSphere Profile-Driven Storage Service.
Hi
Welcome to the communities.
I think its not sufficient to analysis or comment anything.
It would be great if you can upload log file.
"When you fail to plan, you plan to fail."![http://imagicon.info/cat/5-59/vbsmile.png]()
↧
Re: Getting the Virtual Machine Swapfile location using PowerCLI script
I know this is an old post but it proved useful to me the other day. As a result I created a function called Get-VMSwapInfo which was inspired by LucD's response that will output the VM's name, the name of it's swap file, the datastore it's located in, and how much space is left on that datastore. You can get the source and more info on my blog on this page. http://www.virtual-matt.net/2013/06/get-vm-swap-information/
↧
Re: LinkedIn Group for CloudCred
Excelent work Nikki. I've join the CloudCred group on LinkedIn. My username in CloudCred is rhpsystems.
↧
Integrity Check Status -> Out of Date
hi guys
It's been for a week or so my integrity check status is Out of Data.
I try manually running it and it fails. All VDP services are up and running, VDP is at 86% storage utilization.
Something else I have no idea VDP stops sending emails since 3 days ago. Manual email testing works but the automatic one is not arriving.
Any idea how to fix this Integrity Check Status issue?
thanks a lot
↧
↧
Re: Can not Complete the configuration of HA agent on the host. Other HA configuration error
Hi Andre,
Same error comes even after configuring the DNS. I have taken the screen shot with related Event. May be this can be helpful to analyse the issue.
Thanks
Gaurav
↧
vCenter Server not sending data to vCenter Log Insight
I've got a second vCenter server in a different site, the connection tests fine in Log Insight but I'm not seeing any data from it...
The ESXi hosts connected to that vcenter server are reporting their data after running the configure-esxi command from the shell.
Where can I look on the vCenter box thats not reporting to make sure its setup correctly to send the data?
↧
Re: Seperate Cluster for single and dual core VMs
Thanks for the article.. This situation is actually a case of 4 hosts with 150 vms. Of which 50 are dual vcore.. Extremely under powered with ready times over 100k ms most of the time. I'm thinking of separating the dual and single core vms to optimize what little hardware is available, but not sure if it would be worth the trouble. Separate cluster may not be required... I can just setup host preferences on the dual and single core vms to achieve the same distribution.
↧
More Pages to Explore .....
