Thursday, August 08, 2019

vSAN Capacity planning - Understanding vSAN memory consumption in ESXi

It is very clear that VMware vSAN (VMware's software-defined storage) has the momentum in the field, as almost all my customers are planning and designing vSAN in their environments. Capacity planning is an important part of any logical design, so we have to do the same for vSAN. Capacity planning is nothing else than simple math, however, we need to know how the designed system works and what overheads we have to include in our capacity planning exercise. Over the years, a lot of VMware vSphere technical designers did get knowledge and practice how to do capacity planning for core vSphere because server virtualization is here for ages (13+ years). But we all are just starting (3+ years) with vSAN designs, therefore it will take some time to gain the practice and know-how what is important to calculate in terms of VMware hyper-converged infrastructure (VMware HCI = vSphere + vSAN).

One of the many important factors for HCI capacity planning is vSAN memory consumption from the ESXi host memory. There is very good VMware KB 2113954 explaining the math calculation and formulas behind the scene. However, we are tech geeks, so we do not want to do the math on the paper so here is the link to Google Sheets calculator I have prepared for vSAN (All-Flash) memory overhead calculation.

Here is the calculator embedded into this blog post, however, it is just in read-only mode. If you want to change parameters (yellow cells) you have to open Google Sheet available on this link.

Note: I did not finish the calculator for Hybrid configuration because I personally believe that 99% of vSAN deployments should be All-Flash. The reason for this assumption is the fact, that Flash capacity is only 2x or 3x more expensive than magnetic disks and the lower price of magnetic disks is not worth to low speed you can achieve by magnetic disks. In terms of capacity, advanced technics like erasure coding (RAID-5, RAID-6) and deduplication + compression can give you back capacity in All-Flash vSAN as these technics are available or make sense only on All-Flash vSAN. If you would like the same calculator for Hybrid vSAN, leave the comment below this blog post and I will try to find some spare time to prepare another sheet for Hybrid vSAN.

Let's document here some design scenarios with vSAN memory consumptions.

Scenario 1
ESXi host system memory: 192 GB
Number of disk groups: 1
Cache disk size in each disk group: 400 GB
Number of capacity disks in each disk group: 4
vSAN memory consumption per ESXi host is 17.78 GB.

Scenario 2
ESXi host system memory: 192 GB
Number of disk groups: 2
Cache disk size in each disk group: 400 GB
Number of capacity disks in each disk group: 2
vSAN memory consumption per ESXi host is 28 GB.

Scenario 3
ESXi host system memory: 256 GB
Number of disk groups: 2
Cache disk size in each disk group: 400 GB
Number of capacity disks in each disk group: 2
vSAN memory consumption per ESXi host is 28.64 GB.

Hope this is informative and it helps broader VMware community. 

Friday, August 02, 2019

Updating Firmware in vSAN Clusters from VUM

If you operate vSAN you know that correct firmware and drivers are super important for system stability as vSAN software heavily depends on IO controller and physical disks within the server.

Different server vendors have different system management. Some are more complex than other but typical vSphere admin is using vSphere Update Manager (VUM) so would not it be cool to do firmware management directly from VUM? Yes, of course, so how it could be done?

Well, there is a chapter in vSphere documentation covering "Updating Firmware in vSAN Clusters", however, there is no information what hardware is supported. I did some internal research and colleagues have pointed me to VMware KB 60382 "IO Controllers that vSAN supports firmware updating" where supported IO Controllers are listed.

At the time of writing this post, there are plenty of Dell IO Controllers, two Lenovo, two SuperMicro and one Fujitsu.

At the moment, you can update only IO Controllers but it may or may not change in the future.

Wednesday, July 03, 2019

VMware Skyline

VMware Skyline is a relatively new Phone Call or Home Call functionality developed by VMware Global Services. It is a proactive support technology available to customers with an active Production Support or Premier Services contract. Skyline automatically and securely collects, aggregates and analyzes customer specific product usage data to proactively identify potential issues and improve time-to-resolution.

You are probably interested in Skyline Collector System Requirements which are documented here.

Skyline is packaged as a VMware Virtual Appliance (OVA) which is easy to install and operate. From a networking standpoint, there are only two external network connections you have to allow from your environment:
  • HTTPS (443) to
  • HTTPS (443) to
Do you have more questions about Skyline? Your questions can be addressed in Skyline FAQ.

Tuesday, July 02, 2019

vSAN logical design and SSD versus NVMe considerations

I'm just preparing vSAN capacity planning for PoC of one of my customers. Capacity planning for traditional and hyper-converged infrastructure is principally the same. You have to understand TOTAL REQUIRED CAPACITY of your workloads and  USABLE CAPACITY of vSphere Cluster you are designing. Of course, you need to understand how vSAN hyper-converged system conceptually and logically works but it is not rocket science. vSAN is conceptually very straight forward and you can design very different storage systems from performance and capacity point of view. It is just a matter of components you will use. You probably understand that performance characteristics differ if you use rotational SATA disks, SSD or NVMe. For NVMe, 10Gb network can be the bottleneck so you should consider 25Gb network or even more. So, in the figure below is an example of my particular vSAN capacity planning and proposed logical specifications.

Capacity planning is the part of the logical design phase, therefore any physical specifications and details should be avoided. However, within the logical design, you should compare multiple options having an impact on infrastructure design qualities such as

  • availability, 
  • manageability, 
  • scalability, 
  • performance, 
  • security, 
  • recoverability 
  • and last but not least the cost.  

For such considerations, you have to understand the characteristics of different "materials" your system will be eventually built from. When we are talking about magnetic disks, SSD, NVMe, NICs, etc. we are thinking about logical components. So I was just considering the difference between SAS SSD and NVMe Flash for the intended storage system. Of course, different physical models will behave differently but hey, we are in the logical design phase so we need at least some theoretical estimations. We will see the real behavior and performance characteristics after the system is built and tested before production usage or we can invest some time into PoC and validate our expectations.

Nevertheless, cost and performance is always a hot topic when talking with technical architects. Of course, higher performance costs more. However, I was curious about the current situation on the market so I quickly checked the price of SSD and NVMe on e-shop.

Note that this is just the indicative, kind of street price, but it has some informational value.

This is what I have found there today

  • Dell 6.4TB, NVMe, Mixed Use Express Flash, HHHL AIC, PM1725b, DIB - 213,150 CZK
  • Dell 3.84TB SSD vSAS Mixed Use 12Gbps 512e 2.5in Hot-Plug AG drive,3 DWPD 21024 TBW - 105,878 CZK

1 TB of NVMe storage costs 33,281 CZK
1 TB of SAS SSD storage costs 27,572 CZK
This is approximately 20% difference advantage for SSD.

So here are SSD advantages

  • ~ 20% less expensive material
  • scalability because you can put 24 and more SSD disks to 2U rack server but the same server supports usually less than 8 PCIe slots
  • manageability as you can more easily replace disks than PCI cards

The NVMe advantage is the performance with a positive impact on storage latency as SAS SSD has ~250 μs latency and NVMe ~= 80 μs so you should improve performance and storage service quality by a factor of 3.

So as always, you have to consider what infrastructure design quality is good for your particular use case and non-functional requirements and do the right design decision(s) with justification(s).

Any comment? Real experience? Please, leave the comment below the article. 

Monday, June 10, 2019

How to show HBA/NIC driver version

How to find the version of HBA or NIC driver on VMware ESXi?

Let's start with HBA drivers. 

STEP 1/ Find driver name for the particular HBA. In this example, we are interested in vmhba3.

We can use following esxcli command to see driver names ...
esxcli storage core adapter list

So now we have driver name for vmhba3, which is qlnativefc

STEP 2/ Find the driver version.
The following command will show you the version.
vmkload_mod -s qlnativefc | grep -i version

NIC drivers

The process to get NIC driver version is very similar.

STEP 1/ Find driver name for the particular NIC. In this example, we are interested in vmhba3.
esxcli network nic list

STEP 2/ Find the driver version.
The following command will show you the version.
vmkload_mod -s ntg3 | grep -i version

You should always verify your driver versions are at VMware Compatibility Guide. The process of how to do it is documented here How to check I/O device on VMware HCL.

For further information see VMware KB - Determining Network/Storage firmware and driver version in ESXi 4.x and later (1027206)

Thursday, June 06, 2019

vMotion multi-threading and other tuning settings

When you need to boost overall vMotion throughput, you can leverage Multi-NIC vMotion. This is good when you have multiple NICs so it is kind of scale-out solution. But what if you have 40 Gb NICs and you would like to do scale-up and leverage the huge NIC bandwidth (40 Gb) for vMotion?

vMotion is by default using a single thread (aka stream), therefore it does not have enough CPU performance to transfer more than 10 Gb of network traffic. If you really want to use higher NIC bandwidth, the only way is to increase the number of threads pushing the data through the NIC. This is where advanced setting Migrate.VMotionStreamHelpers comes in to play.

I have been informed about these advanced settings by one VMware customer who saw it on some VMworld presentation. I did not find anything in VMware documentation, therefore these settings are undocumented and you should use it with special care.

Advanced System Settings
Number of helpers to allocate for VMotion streams
Max TX queue load (in thousand packet per second) to allow packing on the corresponding RX queue
Threshold (in thousand packet per second) for TX queue load to trigger unpacking of the corresponding RX queue
Maximum length of the Tx queue for the physical NICs 

Wednesday, June 05, 2019

How to get more IOPS from a single VM?

Yesterday, I have got a typical storage performance question. Here is the question ...
I am running a test with my customer how many IOPS we can get from a single VM working with HDS all flash array. The best that I could get with IOmeter was 32K IOPS with 3ms latency at 8KB blocks. No matter what other block size I choose or outstanding IOs, I am unable to have more then 32k. On the other hand I can't find any bottlenecks across the paths or storage. I use PVSCSI storage controller. Latency and queues looks to be ok
IOmeter is good storage test tool. However, you have to understand basic storage principles to plan and interpret your storage performance test properly. The storage is the most crucial component for any vSphere infrastructure, therefore I have some experience with IOmeter and storage performance tests in general and here are my thoughts about this question.

First thing first, every shared storage system requires specific I/O scheduling to NOT give the whole performance to a single worker. The storage worker is the compute process or thread sending storage I/Os down the storage subsystem. If you think about it, it makes a perfect sense as it mitigates the problem of a noisy neighbor. When you invest a lot of money to a shared storage system, you most probably want to use it for multiple servers, right? Does not matter if these servers are physical (ESXi hosts) or virtual (VMs). To get the most performance from shared storage you must use multiple workers and optimally spread them across multiple servers and multiple storage devices (aka LUNs, volumes,  datastores).

IOmeter allows you to use

  • Multiple workers on a single server (aka Manager)
  • Outstanding I/Os within a single worker (asynchronous I/O to a disk queue without waiting for acknowledge)
  • Multiple Managers – the manager is the server generating storage workload (multiple workers) and reporting results to a central IOmeter GUI. This is where IOmeter dynamos come in to play.
To test the performance limits of a shared storage subsystem, it is an always good idea to use multiple servers (IOmeter managers) with multiple workers on each server (nowadays usually VMs) spread across multiple storage devices (datastores / LUNs). This will give you multiple storage queues, which means more parallel I/Os. Parallelism is the way which will give you more performance when such performance exists on shared storage. If such performance does not exist on the shared storage, queueing will not help you to boost performance. If you want, you can also leverage Oustanding I/Os to fill disk queue(s) more quickly and make an additional pressure to a storage subsystem, but it is not necessary if you use the number of workers equal to available queue depth. Outstanding I/Os can help you potentially generating more I/Os with fewer workers but it does not help you to get more performance when your queues are full. You will just increase response times without any positive performance gain.

Just as an example of IOmeter performance test, on the image below, you can see the results from IOmeter distributed performance tests on 2-node vSAN I planned, designed, implemented and tested recently for one of my customers. There is just one disk group (1xSSD cache, 4xSSD capacity).

Above storage performance test was using 8xVMs and each VMs was running 8 storage workers.
I have performed different storage patterns (I/O size, R/W ratio, 100% random access). The performance is pretty good, right? However, I would not be able to get such performance from the single VM having a single vDisk. 
Note: vSAN has a significant advantage in comparison to traditional storage because you do not need to deal with LUNs queueing (HBA Device Queue Depth) as there are no LUNs. On the other hand, in vSAN storage, you have to think about the total performance available for a single vDisk and it boils down to vSAN DiskGroup(s) layout and vDisk object components distribution across physical disks. But that's another topic as the asker is using traditional storage with LUNs.

Unfortunately, using multiple VMs is not the solution for the asker as he is trying to get all I/Os from a single VM.

In the question is declared that a single VM cannot get more than 32K IOPS and observed I/O response time is 3ms. The asker is curious why he cannot get more IOPS from the single VM?

Well, there can be multiple reasons but let’s assume the physical storage is capable provide more than 32K IOPS. I think, that more IOPS cannot be achieved because only one VM is used and IOmeter is using a single vDisk having a single queue. The situation is depicted in drawing below.

So, let’s do the simple math calculation for this particular situation …
  • We have a single vDisk queue having default queue depth 64 (we use Paravirtual SCSI adapter. Non-paravirtualized SCSI adapters have queue depth 32)
  • We have an HBA QLogic having queue default depth 64 (other HBA vendors like Emulex, have default queue depth 32, so it would be another bottleneck on the storage path)
  • The storage has average service time (response time) around 3ms
We have to understand the following basic principles
  • IOPS is the number of I/O operations per second
  • 64 queue depth = 64 I/O operations in parallel = 64 slices for I/O operations
  • Each I/O from these 64 I/Os are in the vDisk queue until SCSI response from the LUN will come back
  • All other I/Os have to wait until there is the free I/O slice in the queue.
And here is the math calculation ...

Q1: How many I/Os can be delivered in this situation per 1 millisecond?
A1: 64 (queue depth) / 3 (service time in ms)  = 64 / 3 = 21.33333 I/Os per 1 millisecond
Q2: How many I/Os can be delivered per 1 second?
A2: It is easy. 1,000 times more than in millisecond. So, 21.33333 x 1,000 = 21333.33 IOs per second ~= 21.3K IOPS
The asker is claiming he can get 32K IOPS with 3 ms response time, therefore it seems that the response time from storage is better than 3 ms. The math above would tell me that storage response time in this particular exercise is somewhere around 2 ms. There can be other mechanisms to boost performance. For example, I/O coalescing but let's keep it simple.

If the storage would be able to service I/O in 1 ms we would be able to get ~64K IOPS.
If the storage would be able to service I/O in 2 ms we would be able to get ~32K IOPS. 
If the storage would be able to service I/O in 3 ms we would be able to get ~21K IOPS. 

The math above would work if END-2-END queue depth is 64. This would be the case when QLogic HBA is used as it has HBA LUN Queue Depth 64. In the case of Emulex HBA, there is HBA LUN Queue Depth 32, therefore higher vDisk Queue Depth (64), would not help.
Hope the principle is clear now.

So how can I boost storage performance for a single VM? If you really need to get more IOPS from the single VM you have only three following options:
  1. increase queue depth, but not only on vDISK itself but END-2-END. IT IS GENERALLY NOT RECOMMENDED as you really must know what you are doing and it can have a negative impact on overall shared storage. However, if you need it and have the justification for it, you can try to tune the system.
  2. use the storage system with low service time (response time). For example, the sub-millisecond storage system (for example 0.5 ms) will give you more IOPS for the same queue depth as a storage system having higher service time (for example 3 ms).
  3. leverage multiple vDisks spread across multiple vSCSI controllers and datastores (LUNs). This would give you more (total) queue depth in a distributed fashion. However, this would have additional requirements for your real application as it would need a filesystem or other mechanism supporting multiple storage devices (vDisks).
I hope options 1 and 2 are clear. Option 3 is depicted in the figure below.

On a typical VMware vSphere environment, you use the shared storage system from multiple ESXi hosts, multiple VMs having vDisks on multiple datastores (LUNs). That's the reason why the default queue depth usually makes perfect sense as it provides fairness among all storage consumers. If you have storage system with, let's say 2 ms response time, and queue depth 32, you can still get around 16K IOPS. This should be good enough for any typical enterprise application, and usually, I recommend to use IOPS limiting to limit some VMs (vDisks) even more. This is how storage performance tiering can be very simply achieved on VMware SDDC with unified infrastructure.  If you need higher storage performance, your application is specific and you should do a specific design and leverage specific technologies or tunings.

By the way, I like Howard's Marks (@DeepStorageNet) statement I have heard on his storage technologies related podcast "GrayBeards".  It is something like ...
"There are only two storage performance types - good enough and not good enough." 
This is very true.
Hope this writeup helps to broader VMware community.

Relevant articles: