After importing the new vSphere Management Assistant (vMA) 4.1 on my VMware Workstation 7 environment I noticed that it boots with a memory crash error, an “unable to access resume device” message and eventually ends with a kernel panic.
The memory crash error seems like something that can be ignored and which is covered by several blog articles about CentOS. The reason why the vMA is locking up is due to the fact that its SCSI card and OS type needs to be changed in the .vmx file
Follow the steps below (credits to Justin for pointing this out to me!)
- Remove the vMA from the VMware Workstation Inventory
- Open the .vmx file and change the following to lines to be:
scsi0.virtualDev = “lsilogic”
guestOS = “rhel5-64″
- Add the vMA back to VMware Workstation and boot it up.
Scripted installation is a new feature for ESXi which is introduced with the release of vSphere 4.1. In this article I’m describing the setup of my home lab in which I use this new feature. Please be advised that this article assumes some basic understanding of Windows and ESX and gives you some guidelines which you can alter for your own specific needs.
There are a few ingredients we need to get this ESXi Scripted Installation to work:
- ESXi Source Repository (HTTP in this article)
- PXELinux
- TFTP Server
- DHCP Server
Read the full article »
Lately I visited a customer who, depending on their customers SLA, provided either SRM protected VM’s (SLA-1) or non-SRM protected VM’s (SLA-2). The technical difference between these SLA’s is that SRM protected VM’s must be placed on LUN’s that are replicated between storage arrays. Different departments are responsible for creating either SLA-1 or SLA-2 VM’s and operational procedures are in place to make sure that the right VMFS datastores is selected while provisioning new VM’s.
Unfortunately VM’s get provisioned on the wrong LUN/VMFS Datastore very often (so either replicated or non-replicated) and this becomes visible in SRM since “Not Configured VM’s” started popping-up as shown in the screenshot below.
Just as a side note, whenever a VM gets provisioned on a datastore that is part of a Protection Group (Datastore Group), the VM gets added to the Protection Group automatically but still needs configuration in order to really get protected by SRM.
The solution is very easy. Since two different departments are provisioning the VM’s it can all be arranged by Datastore Permissions, a feature introduced in vSphere.
In my test environment I created a simple setup. From within the Datastore-view I created a “Non Replicated” and a “Replicated” folder and placed the correct datastores in the folders. Next step is to assign user permissions to the folders, in my example I created a user “test1″ with Administrator permission on the whole environment except on the “Replicated Datastores”-folder.
Whenever user “test1″ wants to provision a VM and store it on a datastore that is placed in the “Replicated Datastores”-folder he will get an error message stating that he is not allowed to.
Please be aware that Datastore Permissions are only valid when connecting via vCenter because it is a vCenter property and only vCenter will check these settings, no changes are made to the volume itself.
Let’s directly start with the most important news; I’ve passed my VMware VCDX defend and got VCDX Number #040! Very Cool!!
The story: two weeks ago I travelled to Munich to do the final and biggest step in the VCDX Certification path – Present and Defend your Design & Implementation Plan.
In preparation to this I worked overtime to first get my design translated and secondly to review all the design choices since you will be ask about this.
The Defend consists of three parts which are covered in the 3-hour timeslot:
1) Defend your submitted design (15 minutes for the presentation, 60 minutes to answer questions about the design)
2) Design Scenario (30 minutes)
3) Troubleshoot Scenario (15 minutes)
I won’t bore you with all the stressy details, I just want to give you some useful tips for the Defend:
- It’s a good idea to put all your diagrams into the slide deck. If there are diagrams that aren’t relevant for the 15 minutes presentation (eg. detailed diagrams) insert them after the “Questions ?”-slide for use as a quick reference whenever questions are fired off;
- Keep focus on the question that is asked you. Don’t make a long story of a simple answer since the defend is time limited and it’s preferable that you score on every subject;
- Make sure you can validate every single decision, know which other (technical) options are available and more importantly what the impact would have been for the other (technical) options;
- Remember that an answer like “I did this since it’s a best practice” isn’t a complete answer, most of the time there must be either a technical and/or organizational reason why a specific choice is made;
- Don’t expect much feedback of the panel members that indicate if the answer you are given is of any use. Most of the time they just start typing and go on to the next question;
- Read Duncan’s recently published blog about the Design and Troubleshoot Scenario and note that the panel members will also tell you that you have to think out loud, ask questions and whiteboard everything that comes up into you mind.
To conclude, keep in mind that VMware published an online directory in which you can find all the VCDX certified people. Remember this resource since I already got in touch with people that dishonestly were claiming that they received their VCDX status!
Lately a customer asked me about Microsoft Cluster Service support on a vSphere iSCSI environment.
The VMware Documentation for vSphere and vSphere Update 1 states:
Before you set up MSCS, review the list of functionality that is not supported for this release, and any requirements and recommendations that apply to your configuration. The following environments and functionality are not supported for MSCS setups with this release of vSphere:
- Clustering on iSCSI, FCoE, and NFS disks.
It wasn’t clear however if this also affects a Microsoft Cluster that is created with an iSCSI initiator running inside the guest OS. I’ve contacted VMware Support on this case and they told me:
We don’t support iscsi for mscs even if the software initiator in run from within windows, or with any other method. Technically you might be able to get it to work but you will have an unsupported setup if you ever hit any issues. As such I cannot recommend you go down this route.
So we can conclude that within vSphere, Microsoft Cluster Service (Failover Clustering) isn’t supported on iSCSI, FCoE and NFS no matter what the configuration looks like.
Its official now, I’m joining VMware PSO starting at the 1st of April! (no joke) and will become a direct colleague of Frank Denneman who joined PSO earlier this year.
Last year I’ve spoken to Duncan Epping about PSO and the fact that he was thinking about a change in his position. His career move obviously meant that a PSO position will be available and recently VMware and I came to an agreement. It’s needless to say that I’m very excited for this job role and that I’m going to make the best of it!
Last but not least I want to thank my current employer inPeople for their support and understanding. Thanks guys!
by admin on March 4, 2010
I recently got a question from the backup architect asking me about LeftHand snapshot techniques to make backups of VM’s. The current storage design classifies VM’s based on their Recovery Point Objective (RPO) per LUN as shown in the overview below.
Within this storage design it could be a good solution if we automate LeftHand Snapshots bases on the LUN classification. So for instance we could create a schedule that creates a LeftHand Snapshot every 4 hours for volume “VMFS_R4_01” and every 12 hours for volume “VMFS_R12_01”.
Read the full article »
Everyone knows the situation in where a (s)VMotion asks you to select a Resource Pool. I would love to see that VMware automatically selects the current VM’s Resource Pool but unfortunately we have to make a selection since it defaults to the cluster. When we leave it to default it obviously causes the VM to get moved out of it’s Resource Pool.
Follow the instruction below to get the current VM’s Resource Pool without having to walk thru each Resource Pool manually.
1) Select the “Host and Clusters”-view
2) Type the VM name in the “Search Field”, press enter and select the VM
3) This will unfold the correct Resource Pool on the left side of the screen and gives you just the information you need when doing a (s)VMotion.
Lately I’ve been asked to review a PID (Project Initiation Document) for a new File Server project based on Linux/GFS v2 (General File System) hosted on VMware vSphere. I noticed that the PID was talking about a dynamic File Server that could (and even would) grow towards 100 TB storage space. That number immediately set off alarm bells in my head.
The vSphere Configuration Maximums document states that a VM can have a maximum of 4 SCSI cards.
So my initial design would look like this:
- 1 SCSI Card is for the VM Boot Disk (SCSI Bus Sharing: None) which leaves us with
- 3 SCSI Cards for the RDM’s (SCSI Bus Sharing: Virtual or Physical).
Read the full article »
After rebooting several Virtual Connect Modules to test the failover behaviors I got myself in a situation in were the Virtual Connect Manager got completely unresponsive. In my case the vcutil from HP eventually solved my problem so I want to give some more information on this tool since I only knew it as a Virtual Connect Firmware Update Tool from the past.
In my case the following statements were true:
- While logging on with the web browser the Interface the “Loading, please wait…” wouldn’t disappear.
- While logging on with SSH, I was able to enter credentials but after that the CLI never appeared and eventually timed out.
Read the full article »
Subscribe
