Yellow Bricks

by Duncan Epping

Storage

vSAN File Services and Stretched Clusters!

· · Leave a Comment

As most of you probably know, vSAN File Services is not supported on a stretched cluster with vSAN 7.0 or 7.0U1. However, starting with vSAN 7.0 U2 we now fully support the use of vSAN File Services on a stretched cluster configuration! Why is that?

In 7.0 U2, you now have the ability to specify during configuration of vSAN File Services to which site certain IP addresses belong. In other words, you can specify the “site affinity” of your File Service services. This is shown in the screenshot below. Now I do want to note, this is a soft affinity rule. Meaning that if hosts, or VMs, fail on which these file services containers are running it could be that the container is restarted in the opposite location. Again, a soft rule, not a hard rule!

Of course, that is not the end of the story. You also need to be able to specify for each share with which location they have affinity. Again, you can do this during configuration (or edit it afterward if desired), and this basically then sets the affinity for the file share to a location. Or said differently, it will ensure that when you connect to file share, one of the file servers in the specified site will be used. Again, this is a soft rule, meaning that if none of the file servers are available on that site, you will still be able to use vSAN File Services,  just not with the optimized data path you defined.

Hopefully, that gives a quick overview of how you can use vSAN File Services in combination with a vSAN Stretched Cluster.  I created a video to demonstrate these new capabilities, you can watch it below.

vSAN 7.0 U2 now integrates with vSphere DRS

· · 1 Comment

One of the features our team requested a while back was integration between DRS and vSAN. The key use case we had was for stretched clusters. Especially in scenarios where a failure has occurred, it would be useful if DRS would understand what vSAN is doing. What do I mean by that?

Today when customers create a stretched cluster they have two locations. Using vSAN terminology these locations are referred to as the Preferred Fault Domain and the Secondary Fault Domain. Typically when VMs are then deployed, customers will create VM-to-Host Affinity Rules which state that VMs should reside in a particular location. When these rules are created DRS will do its best to ensure that the defined rule is adhered to. What is the problem?

Well if you are running a stretched cluster and let’s say one of the sites go down, then what happens when the failed location returns for duty is the following:

  • vSAN detects the missing components are available again
  • vSAN will start the resynchronization of the components
  • DRS runs every minute and rebalances and will move VMs based on the DRS rules

This means that the VMs for which rules are defined will move back to their respective location, even though vSAN is potentially still resynchronizing the data. First of all, the migration will interfere with the replication traffic. Secondly, for as long as the resync has not completed, I/O will across the network between the two locations, this will not only interfere with resync traffic, it will also increase latency for those workloads. So, how does vSAN 7.0 U2 solve this?

Starting with vSAN 7.0 U2 and vSphere 7.0 U2 we now have DRS and vSAN communicating. DRS will verify with vSAN what the state is of the environment, and it will not migrate the VMs back as long the VMs are healthy again. When the VMs are healthy and the resync has completed, you will see the rules being applied and the VMs automatically migrate back (when DRS is configured to Fully Automated that is).

I can’t really show it with a screenshot or anything, as this is a change in the vSAN/DRS architecture, but to make sure it worked I recorded a quick demo which I published through Youtube. Make sure to watch the video!

vSAN 7.0 U2 Durability Components?

· · Leave a Comment

Last week I published a new demo on my youtube channel (at the bottom of this post) and it discussed an enhanced feature called Durability Components. Some may know these as “delta components” as well. These durability components were introduced in vSAN 7.0 Update 1 and provided a mechanism to maintain the required availability for VMs while doing maintenance. That meaning that when you would place a host into maintenance mode new “durability components” would be created for the components which were stored on that host. This would then allow all the new VM I/O to be committed to the existing component, as well as the durability component.

Now, starting with vSAN 7.0 Update 2, vSAN also uses these durability components in situations where a host failure has occurred. So if a host has failed, durability components will be created to ensure we still maintain the specified availability level specified within the policy as shown in the diagram above. The great thing is that if a second host fails in an FTT=1 scenario and you are able to recover the first failed host, we can still merge the data with the first failed host with the durability component! So not only are these durability components great for improving the resync times, but they also provide a higher level of availability to vSAN! To summarize:

  1. Host fails
  2. Durability components are created for all impacted objects
  3. New writes are committed to existing components and the new durability components
  4. Host recovers
  5. Durability components are merged with the previously failed components
  6. Durability components are deleted when resync has completed

I hope that help providing a better understanding of how these durability components help improving availability/resiliency in your environment with vSAN 7.0 Update 2.

I can understand that some of you may not want to test durability components in their own environment, this is why I recorded a quick demo and published it on my youtube channel. Check out the video below, as it also shows you how durability components are represented in the UI.

vSAN 7.0 U2 Skyline Health History

· · Leave a Comment

I have already discussed this briefly during my 7.0 U2 video presentation, which can be found here, but I also wanted to share the demo I recorded with you, and provide some additional details. Over the past years, one of the most requested features for Skyline Health, or Health Check as it used to be called, was the ability to go back in time to see what has happened between certain dates. This functionality was demonstrated at VMworld, and a few VMUGs the past year, and now finally made it into the release with vSAN 7.0 U2.

The “health history” feature simply provides a toggle that enables you to go back in time. When you tick the toggle, you can specify a date range. Note that the range can be anywhere between the current date, and 30 days back. The health check data is stored, for 30 days, in the vCenter Server database. This is important to know because if you feel that there’s no need to store the data, you can disable the feature under vSAN Services in the Configuration tab. When you disable the feature the data is deleted from the vCenter Server database. Now if you flip the toggle, set a date range, and look at the different checks you should see green checks. If a check is not green, but rather red or orange you can click the check.

When clicking on the red square, you will see which check failed, and when exactly it failed. The number in the square or circle refers to the number of checks that were run and resulted in the same state. In other words, 37 red, 45 green, 54 red checks etc. When you click on it, you will see the below.

Hopefully, this will then allow you, during troubleshooting, to correlate particular failures or configuration changes, to the issue that bubbled up in the health check. I feel that having the date/time is already very useful, as it will allow you to focus on a more specific date/time range while reading logs or going through the events section.

Anyway, if you would like to see the feature in action, check out the demo below.

VMs which are not stretched in a stretched cluster, which policy to use?

· · 2 Comments

I’ve seen this question popping up regularly. Which policy setting (“site disaster tolerance” and “failures to tolerate”) should I use when I do not want to stretch my VMs? Well, that is actually pretty straight forward, in my opinion, you really only have two options you should ever use:

  • None – Keep data on preferred (stretched cluster)
  • None – Keep data on non-preferred (stretched cluster)

Yes, there is another option. This option is called “None – Stretched Cluster” and then there’s also “None – Standard Cluster”. Why should you not use these? Well, let’s start with “None – Stretched Cluster”. In the case of “None – Stretched Cluster”, vSAN will per object decide where to place it. As you hopefully know, a VM consists of multiple objects. As you can imagine, this is not optimal from a performance point of view, as you could end up having a VMDK being placed in Site A and a VMDK being placed in Site B. Which means it would read and write from both locations from a storage point of view, while the VM would be sitting in a single location from a compute point of view. It is also not very optimal from an availability stance, as it would mean that when the intersite link is unavailable, some objects of the VM would also become inaccessible. Not a great situation. What would it look like? Well, potentially something like the below diagram!

Then there’s “None – Standard Cluster”, what happens in this case? When you use “None – Standard Cluster” with “RAID-1”, what is going to happen is that the VM is configured with FTT=1 and RAID-1, but in a stretched cluster “FTT” does not exist, and FTT automatically will become PFTT. This means that the VM is going to be mirrored across locations, and you will have SFTT=0, which means no resiliency locally. It is the same as “Dual Site Mirroring”+”No Data Redundancy”!

In summary, if you ask me, “none – standard cluster” and “none – stretched cluster” should not be used in a stretched cluster.