NVMe Storage Server Project

 End User Computing, ESXi, Homelab, HPE, IOCREST, iSCSI, NFS, NVMe, SSD, Storage, TrueNAS, VDI, VMware, vSphere  12 Responses »
May 012021
 
Picture of NVMe Storage Server Project

For over a year and a half I have been working on building a custom NVMe Storage Server for my homelab. I wanted to build a high speed storage system similar to a NAS or SAN, backed with NVMe drives that provides iSCSI, NFS, and SMB Windows File Shares to my network.

The computers accessing the NVMe Storage Server would include VMware ESXi hosts, Raspberry Pi SBCs, and of course Windows Computers and Workstations.

The focus of this project is on high throughput (in the GB/sec) and IOPS.

The current plan for the storage environment is for video editing, as well as VDI VM storage. This can and will change as the project progresses.

The History

More and more businesses are using all-flash NVMe and SSD based storage systems, so I figured there’s no reason why I can’t have build and have my own budget custom all NVMe flash NAS.

This is the story of how I built my own NVMe based Storage Server.

The first version of the NVMe Storage Server consisted of the IO-PEX40152 card with 4 x 2TB Sabrent Rocket 4 NVMe drives inside of an HPE Proliant DL360p Gen8 Server. The server was running ESXi with TrueNAS virtualized, and the PCIe card passed through to the TrueNAS VM.

The results were great, the performance was amazing, and both servers had access to the NFS export via 2 x 10Gb SFP+ networking.

There were three main problems with this setup:

  1. Virtualized – Once a month I had an ESXi PSOD. This was either due to overheating of the IO-PEX40152 card because of modifications I made, or bugs with the DL360p servers and PCIe passthrough.
  2. NFS instead of iSCSI – Because TrueNAS was virtualized inside of the host that was using it for storage, I had to use NFS since the host virtualizing TrueNAS would also be accessing the data on the TrueNAS VM. When shutting down the host, you need to shut down TrueNAS first. NFS disconnects are handled way healthier than iSCSI disconnects (which can cause corruption even if no files are being used).
  3. CPU Cores maxed on data transfer – When doing initial testing, I was maxing out the CPU cores assigned to the TrueNAS VM because the data transfers were so high. I needed a CPU and setup that was better fit.

Version 1 went great, but you can see some things needed to be changed. I decided to go with a dedicated server, not virtualize TrueNAS, and go for a newer CPU with a higher Ghz speed.

And so, version 2 was born (built). Keep reading and scrolling for pictures!

The Hardware

On version 2 of the project, the hardware includes:

Notes on the Hardware:

  • While the ML310e Gen8 v2 server is a cheap low entry server, it’s been a fantastic team member of my homelab.
  • HPE Dual 10G Port 560SFP+ adapters can be found brand new in unsealed boxes on eBay at very attractive prices. Using HPE Parts inside of HPE Servers, avoids the fans from spinning up fast.
  • The ML310e Gen8 v2 has some issues with passing through PCIe cards to ESXi. Works perfect when not passing through.

The new NVMe Storage Server

I decided to repurpose an HPE Proliant ML310e Gen8 v2 Server. This server was originally acting as my Nvidia Grid K1 VDI server, because it supported large PCIe cards. With the addition of my new AMD S7150 x2 hacked in/on to one of my DL360p Gen8’s, I no longer needed the GRID card in this server and decided to repurpose it.

Picture of an HPe ML310e Gen8 v2 with NVMe Storage
HPe ML310e Gen8 v2 with NVMe Storage

I installed the IOCREST IO-PEX40152 card in to the PCIe 16x slot, with 4 x 2TB Sabrent Rocket 4 NVME drives.

Picture of IOCREST IO-PEX40152 with GLOTRENDS M.2 NVMe SSD Heatsink on Sabrent Rocket 4 NVME
IOCREST IO-PEX40152 with GLOTRENDS M.2 NVMe SSD Heatsink on Sabrent Rocket 4 NVME

While the server has a PCIe 16x wide slot, it only has an 8x bus going to the slot. This means we will have half the capable speed vs the true 16x slot. This however does not pose a problem because we’ll be maxing out the 10Gb NICs long before we max out the 8x bus speed.

Picture of an HPE ML310e Gen8 v2 with IOCREST IO-PEX40152
HPE ML310e Gen8 v2 with IOCREST IO-PEX40152
Lookdown view of HPE ML310e Gen8 v2 with IOCREST IO-PEX40152
HPE ML310e Gen8 v2 with IOCREST IO-PEX40152

I also installed an HPE Dual Port 560SFP+ NIC in to the second slot. This will allow a total of 2 x 10Gb network connections from the server to the Ubiquiti UniFi US-16-XG 10Gb network switch, the backbone of my network.

Picture of HPE ML310e Gen8 v2 with HPE 560SFP+ and 10Gig DAC
HPE ML310e Gen8 v2 with HPE 560SFP+ and 10Gig DAC
Picture of HPE ML310e Gen8 v2 with HPE 560SFP+ and 10Gig DAC
HPE ML310e Gen8 v2 with HPE 560SFP+ and 10Gig DAC

The Server also have 4 x Hot Swappable HD bays on the front. When configured in HBA mode (via the BIOS), these are accessible by TrueNAS and can be used. I plan on populating these with 4 x 4TB HPE MDL SATA Hot Swappable drives to act as a replication destination for the NVMe pool and/or slower magnetic long-term storage.

Front view of HPE ML310e Gen8 v2 with Hotswap Drive bays
HPE ML310e Gen8 v2 with Hotswap Drive bays

I may also try to give WD RED Pro drives a try, but I’m not sure if they will cause the fans to speed up on the server.

TrueNAS Installation and Configuration

For the initial Proof-Of-Concept for version 2, I decided to be quick and dirty and install it to a USB stick. I also waited until I installed TrueNAS on to the USB stick and completed basic configuration before installing the Quad NVMe PCIe card and 10Gb NIC. I’m using a USB 3.0 port on the back of the server for speed, as I can’t verify if the port on the motherboard is USB 2 or USB 3.

Picture of a TrueNAS USB Stick on HPE ML310e Gen8 v2
TrueNAS USB Stick on HPE ML310e Gen8 v2

TrueNAS installation worked without any problems whatsoever on the ML310e. I configured the basic IP, time, accounts, and other generic settings. I then proceeded to install the PCIe cards (storage and networking).

Screenshot of TrueNAS Dashboard Installed on NVMe Storage Server
TrueNAS Installed on NVMe Storage Server

All NVMe drives were recognized, along with the 2 HDDs I had in the front Hot-swap bays (sitting on an HP B120i Controller configured in HBA mode).

Screenshot of available TrueNAS NVMe Disks
TrueNAS NVMe Disks

The 560SFP+ NIC also was detected without any issues and available to configure.

Dashboard Screenshot of TrueNAS 560SFP+ 10Gb NIC
TrueNAS 560SFP+ 10Gb NIC

Storage Configuration

I’ve already done some testing and created a guide on FreeNAS and TrueNAS ZFS Optimizations and Considerations for SSD and NVMe, so I made sure to use what I learned in this version of the project.

I created a striped pool (no redundancy) of all 4 x 2TB NVMe drives. This gave us around 8TB of usable high speed NVMe storage. I also created some datasets and a zVOL for iSCSI.

Screenshot of NVMe TrueNAS Storage Pool with Datasets and zVol
NVMe TrueNAS Storage Pool with Datasets and zVol

I chose to go with the defaults for compression to start with. I will be testing throughput and achievable speeds in the future. You should always test this in every and all custom environments as the results will always vary.

Network Configuration

Initial configuration was done via the 1Gb NIC connection to my main LAN network. I had to change this as the 10Gb NIC will be directly connected to the network backbone and needs to access the LAN and Storage VLANs.

I went ahead and configured a VLAN Interface on VLAN 220 for the Storage network. Connections for iSCSI and NFS will be made on this network as all my ESXi servers have vmknics configured on this VLAN for storage. I also made sure to configure an MTU of 9000 for jumbo frames (packets) to increase performance. Remember that all hosts must have the same MTU to communicate.

Screenshot of 10Gb NIC on Storage VLAN
10Gb NIC on Storage VLAN

Next up, I had to create another VLAN interface for the LAN network. This would be used for management, as well as to provide Windows File Share (SMB/Samba) access to the workstations on the network. We leave the MTU on this adapter as 1500 since that’s what my LAN network is using.

Screenshot of 10Gb NIC on LAN VLAN
10Gb NIC on LAN VLAN

As a note, I had to delete the configuration for the existing management settings (don’t worry, it doesn’t take effect until you hit test) and configure the VLAN interface for my LANs VLAN and IP. I tested the settings, confirmed it was good, and it was all setup.

At this point, only the 10Gb NIC is now being used so I went ahead and disconnected the 1Gb network cable.

Sharing Setup and Configuration

It’s now time to configure the sharing protocols that will be used. As mentioned before, I plan on deploying iSCSI, NFS, and Windows File Shares (SMB/Samba).

iSCSI and NFS Configuration

Normally, for a VMware ESXi virtualization environment, I would always usually prefer iSCSI based storage, however I also wanted to configure NFS to test throughput of both with NVMe flash storage.

Earlier, I created the datasets for all my my NFS exports and a zVOL volume for iSCSI.

Note, that in order to take advantage of the VMware VAAI storage directives (enhancements), you must use a zVOL to present an iSCSI target to an ESXi host.

For NFS, you can simply create a dataset and then export it.

For iSCSI, you need to create a zVol and then configure the iSCSI Target settings and make it available.

SMB (Windows File Shares)

I needed to create a Windows File Share for file based storage from Windows computers. I plan on using the Windows File Share for high-speed storage of files for video editing.

Using the dataset I created earlier, I configured a Windows Share, user accounts, and tested accessing it. Works perfect!

Connecting the host

Connecting the ESXi hosts to the iSCSI targets and the NFS exports is done in the exact same way that you would with any other storage system, so I won’t be including details on that in this post.

We can clearly see the iSCSI target and NFS exports on the ESXi host.

Screenshot of TrueNAS NVMe iSCSI Target on VMware ESXi Host
TrueNAS NVMe iSCSI Target on VMware ESXi Host
Screenshot of NVMe iSCSI and NFS ESXi Datastores
NVMe iSCSI and NFS ESXi Datastores

To access Windows File Shares, we log on and map the network share like you would normally with any file server.

Testing

For testing, I moved (using Storage vMotion) my main VDI desktop to the new NVMe based iSCSI Target LUN on the NVMe Storage Server. After testing iSCSI, I then used Storage vMotion again to move it to the NFS datastore. Please see below for the NVMe storage server speed test results.

Speed Tests

Just to start off, I want to post a screenshot of a few previous benchmarks I compiled when testing and reviewing the Sabrent Rocket 4 NVMe SSD disks installed in my HPE DL360p Gen8 Server and passed through to a VM (Add NVMe capability to an HPE Proliant DL360p Gen8 Server).

Screenshot of CrystalDiskMark testing an IOCREST IO-PEX40152 and Sabrent Rocket 4 NVME SSD for speed
CrystalDiskMark testing an IOCREST IO-PEX40152 and Sabrent Rocket 4 NVME SSD
Screenshot of CrystalDiskMark testing IOPS on an IOCREST IO-PEX40152 and Sabrent Rocket 4 NVME SSD
CrystalDiskMark testing IOPS on an IOCREST IO-PEX40152 and Sabrent Rocket 4 NVME SSD

Note, that when I performed these tests, my CPU was maxed out and limiting the actual throughput. Even then, these are some fairly impressive speeds. Also, these tests were directly testing each NVMe drive individually.

Moving on to the NVMe Storage Server, I decided to test iSCSI NVMe throughput and NFS NVMe throughput.

I opened up CrystalDiskMark and started a generic test, running a 16GB test file a total of 6 times on my VDI VM sitting on the iSCSI NVMe LUN.

Screenshot of NVMe Storage Server iSCSI Benchmark with CrystalDiskMark
NVMe Storage Server iSCSI Benchmark with CrystalDiskMark

You can see some impressive speeds maxing out the 10Gb NIC with crazy performance of the NVME storage:

  • 1196MB/sec READ
  • 1145.28MB/sec WRITE (Maxing out the 10GB NIC)
  • 62,725.10 IOPS READ
  • 42,203.13 IOPS WRITE

Additionally, here’s a screenshot of the ix0 NIC on the TrueNAS system during the speed test benchmark: 1.12 GiB/s.

Screenshot of TrueNAS NVME Maxing out 10Gig NIC
TrueNAS NVME Maxing out 10Gig NIC

And remember this is with compression. I’m really excited to see how I can further tweak and optimize this, and also what increases will come with configuring iSCSI MPIO. I’m also going to try to increase the IOPS to get them closer to what each individual NVMe drive can do.

Now on to NFS, the results were horrible when moving the VM to the NFS Export.

Screenshot of NVMe Storage Server NFS Benchmark with CrystalDiskMark
NVMe Storage Server NFS Benchmark with CrystalDiskMark

You can see that the read speed was impressive, but the write speed was not. This is partly due to how writes are handled with NFS exports.

Clearly iSCSI is the best performing method for ESXi host connectivity to a TrueNAS based NVMe Storage Server. This works perfect because we’ll get the VAAI features (like being able to reclaim space).

iSCSI MPIO Speed Test

This is more of an update… I was finally able to connect, configure, and utilize the 2nd 10Gbe port on the 560SFP+ NIC. In my setup, both hosts and the TrueNAS storage server all have 2 connections to the switch, with 2 VLANs and 2 subnets dedicated to storage. Check out the before/after speed tests with enabling iSCSI MPIO.

Screenshot of TrueNAS NVME iSCSI MPIO Before
TrueNAS NVME iSCSI MPIO Before
Screenshot of TrueNAS NVME iSCSI MPIO After
TrueNAS NVME iSCSI MPIO After
Before and After enabling iSCSI MPIO on TrueNAS with NVME Storage

As you can see I was able to essentially double my read speeds (again maxing out the networking layer), however you’ll notice that the write speeds maxed out at 1598MB/sec. I believe we’ve reached a limitation of the CPU, PCIe bus, or something else inside of the server. Note, that this is not a limitation of the Sabrent Rocket 4 NVME drives, or the IOCREST NVME PCIe card.

Moving Forward

I’ve had this configuration running for around a week now with absolutely no issues, no crashes, and it’s been very stable.

Using a VDI VM on NVMe backed storage is lightning fast and I love the experience.

I plan on running like this for a little while to continue to test the stability of the environment before making more changes and expanding the configuration and usage.

Future Plans (and Configuration)

  • Drive Bays
    • I plan to populate the 4 hot-swappable drive bays with HPE 4TB MDL drives. Configured with RaidZ1, this should give me around 12TB usable storage. I can use this for file storage, backups, replication, and more.
  • NVMe Replication
    • This design was focused on creating non-redundant extremely fast storage. Because I’m limited to a total of 4 NVMe disks in this design, I chose not to use RaidZ and striped the data. If one NVMe drive is lost, all data is lost.
    • I don’t plan on storing anything important, and at this point the storage is only being used for VDI VMs (which are backed up), and Video editing.
    • If I can populate the front drive bays, I can replicate the NVMe storage to the traditional HDD storage on a frequent basis to protect against failure to some level or degree.
  • Version 3 of the NVMe Storage Server
    • More NVMe and Bigger NVMe – I want more storage! I want to test different levels of RaidZ, and connect to the backbone at even faster speeds.
    • NVME Drives with PLP (Power Loss Prevention) for data security and protection.
    • Dual Power Supply

Let me know your thoughts and ideas on this setup!

vCSA 7.0.1 (7.0 U1 Build 16858589) update failed

 vCenter vCSA, VMware, vSphere  No Responses »
Oct 232020
 
vCSA Update Installation

When updating VMware vCenter vCSA 7.0 U1 (Build 16858589) to vCSA 7.0 U1 (Build 17004997/17005016, Version 7.0.1.00100), you may notice that the update fails and reports issues with pre-update checks.

Pre-update checks done prior to the update will pass and allow you to proceed, however it’s the installation that will fail and crash reporting this error.

After the installation fails, you will no longer be able to log in to the vCSA VAMI reporting the error “Unable to Login” using the root account.

You are able to login via SSH. Resetting the root password via SSH will not resolve this issue.

The Problem

In the past, issues with the root password expiring have caused similar behavior on the vCSA VAMI. Changing the root password does not resolve this specific issue.

Further troubleshooting, it appears that special characters in the root password such as “!”, “.”, and “@” caused this issue to occur in my environment.

I was not able to fix the broken vCSA after the failed update. Access to the vCSA was not possible, however vCenter functions were still operating.

The Solution

To resolve this situation in my environment, I restored a snapshot of the vCSA taken prior to updating.

After restoring the snapshot, I changed the root password for VAMI and restarted the vCSA.

Another snapshot was taken prior to attempting the upgrade, which was now succesfull after removing special characters out of the root password.

What is the vCLS Virtual Machine on VMware

 VMware, vSphere  16 Responses »
Oct 152020
 
VMware vCLS VM in VM List

Did a new VM appear on your VMware vSphere cluster called “vCLS”? Maybe multiple appeared named “vCLS (1)”, “vCLS (2)”, “vCLS (3)” appeared.

VMware vCLS VM in vSphere Cluster Objects
VMware vCLS VM in vSphere Cluster Objects

This could be frightening but fear not, this is part of VMware vSphere 7.0 Update 1.

What is the vCLS VM?

The vCLS virtural machine is essentially an “appliance” or “service” VM that allows a vSphere cluster to remain functioning in the event that the vCenter Server becomes unavailable. It will maintain the health and services of that cluster.

Where did the vCLS VM come from?

The vCLS VM will appear after upgrading to vSphere 7.0 Update 1. I’m assuming it was deployed during the upgrade process.

It does not appear in the standard Cluster, Hosts, and VMs view, but does appear when looking at the vSphere objects VM lists, Storage VM lists, etc…

Is it normal to have more than one vCLS VM?

The vCLS VMs are created when hosts are added to a vSphere Cluster. Up to 3 vCLS VMs are required to run in each vSphere Cluster.

The vCLS VMs will also appear on clusters which contain only one or two hosts. These configurations will result in either 1 or 2 vCLS VMs named “vCLS (1)” and “vCLS (2)”.

A note on licensing in regards to the vCLS VM

For VMware environments that use VM based licensing like vSphere for ROBO (Remote Office Branch Office), vCLS VMs are shown in the licensing interface as counting towards licensed VMs. Please Note that these VMs do not official count towards your purchased licenses as these are VMware System VMs. Please read VMware KB 80472 for more information on this.

More Information on vCLS VMs

For more information and technical specifics, you can visit the link below:

https://docs.vmware.com/en/VMware-vSphere/7.0/com.vmware.vsphere.vcenterhost.doc/GUID-96BD6016-4BE7-4B1C-8269-568D1555B08C.html

Hope this posts helps, and puts some minds at ease. Your VMware environment has NOT been compromised.

Troubleshooting vMotion using the vMotion TCP/IP stack

 VMware, vSphere  10 Responses »
May 252020
 
vSphere Logo Image

When troubleshooting connectivity issues with your vMotion network (or vMotion VLAN), you may notice that you’re unable to ping using the ping or vmkping command on your ESXi and VMware hosts.

This occurs when you’re suing the vMotion TCP/IP stack on your vmkernel (vmk) adapters that are configured for vMotion.

This also applies if you’re using long distance vMotion (LDVM).

Why

The vMotion TCP/IP stack requires special syntax for ping and ICMP tests on the vmk adapters.

A screenshot of vmk adapters, one of which is using the vMotion TCP/IP Stack
VMK using vMotion TCP/IP Stack

Above is an example where a vmk adapter (vmk3) is configured to use the vMotion TCP/IP stack.

How

To “ping” and test your vMotion network that uses the vMotion TCP/IP stack, you’ll need to use the special command below:

esxcli network diag ping -I vmk1 --netstack=vmotion -H ip.add.re.ss

In the command above, change “vmk1” to the vmkernel adapter you want to send the pings from. Additionally, change “ip.add.re.ss” to the IP address of the host you want to ping.

Using this method, you can fully verify network connectivity between the vMotion vmks using the vMotion stack.

Additional information and examples can be found at https://kb.vmware.com/s/article/59590.

vSphere Mobile Client for Android Mobile Devices

 VMware, vSphere  1 Response »
Jun 122019
 
VMware vSphere Mobile Watchlist Logo

It’s finally here! VMware has released the alpha (test) of the vSphere Mobile Client for Android Devices. This will allow you to manage your vSphere instance via your Android mobile device.

Some of you may remember the vSphere Mobile Watchlist app for android. It was great because it allowed you to manage your vSphere environment (and I still use it), but one day it was abruptly removed from the Google Play store and no longer available. However, those that had it installed could keep using it.

This new version of the vSphere Mobile Client is only available for Android as of the time of this post.

vSphere Mobile Client Fling

The VMware fling is here: https://labs.vmware.com/flings/vsphere-mobile-client

While there is a tarball download, you’ll actually want to forget that and follow the instructions for a proper install. The tarball is only needed if you want to deploy the notification service.

Installing the vSphere Mobile Client for Android

First, you need to join the alpha testers group here: https://groups.google.com/forum/#!forum/vsphere-mobile-client/join

Second, you need to opt-in to the Google Play Test app here: https://play.google.com/apps/testing/com.vmware.vsphere.cloudsmith

Then simply follow the instruction after the opt-in and download it for your device.

Using the vSphere Mobile Client for Android

The app is a slick but simple one. Since it’s alpha, functionality is limited, but gives you the ability shutdown, restart, view performance and do a couple other things.

Bugs and Annoyances

Shortly after using the app, I noticed that I couldn’t log in subsequent tries due to an “incorrect user name or password”. I know I was typing it right, so I’m assuming this is a bug. To resolve this, you have to delete the app cache, then you will be able to log in properly.

Unfortunately the app also doesn’t allow you to save your password, like the previous watchlist app.

Screenshots

See below for some screenshots:

  • VMware vSphere Mobile Client Login Screen
  • VMware vSphere Mobile Client VM List Screen
  • VMware vSphere Mobile Client Performance Screen

Conclusion

All in all, it’s pretty exciting that VMware is finally working on an official mobile app. I still use watchlist almost daily, so I see tremendous value in this!

Leave a comment below and let me know what you think of the new app!

VMFS Auto UNMAP (Space Reclamation) not working on vSphere

 VMware, vSphere  1 Response »
Nov 172018
 

When running VMware vSphere 6 or vSphere 7 and ESXi on your hosts with VMFS6, you may notice that auto unmap (space reclamation) is not working even though it is enabled. In addition, you’ll find that manual unmap functions still work.

Why is UNMAP not working

This is because your storage array (SAN) may have a larger unmap granularity block size than 1MB. VMFS version 6 (source) requires an unmap granularity of 1MB and does not support automatic unmap on arrays that are larger.

For example, on the HPE MSA 2040 the page size when using virtual storage is 4MB, hence auto unmap is not supported and does not work. You can still manually perform unmap on arrays with block/page sizes larger than 1MB.

Additional Information and Resources

Perform manual VMFS unmap on vSphere 6.5 and 6.7 with VMFS 6 – https://www.stephenwagner.com/2017/02/07/vmfs-unmap-command-vsphere-6-5-with-vmfs-6-auto/

vSphere 6.5, 6.7 and VMFS 6 – Change storage reclaim priority from low to medium or high – https://www.stephenwagner.com/2017/02/08/vsphere-vmfs-6-change-storage-reclaim-priority-low-medium-high/

Release unused space on host and guest filesystems with thin-provisioned Sophos UTM appliance (SW) – https://www.stephenwagner.com/2018/01/18/release-unused-space-vmdk-thin-provisioned-sophos-utm/

VMware vDP (vSphere Data Protection) actively being maintainted for rest of VMware vSphere 6.5 Updates

 vDP, VMware, vSphere  No Responses »
Oct 212018
 

VMware announced the end of availability for the vDP (vSphere Data Protection) product last year. In their release, it was mentioned that no more updates would be issued for the product.

At that time, I recommended to most customers to start planning to upgrade/migrate to a different product. VMware then released some ESXi and vSphere upgrades, but didn’t immediately update the vDP appliance. It appeared as though no more updates would be issued.

Due to the the previous announcement, and a lack of an updated vDP appliance, this increased the urgency of migrating to a different product.

However, just recently I noticed that VMware is still maintaining the vDP appliance for vSphere 6.5, so you can continue to use it while you plan your migration to a new Backup/DR product, and your upgrade to vSphere version 6.7.

 

Please note: I noticed the updates to the vDP appliance are being released later than the vSphere/ESXi updates. I wouldn’t upgrade your ESXi hosts to 6.5 certain update levels until the applicable vDP updated appliance is available for that version.

VMware ESXi asyncUnmapFile grows massive, cannot delete, errors

 ESXi, iSCSI, VMware, vSphere  No Responses »
Oct 202018
 

On an ESXi host when performing a manual unmap on your storage datastore, you may notice a very large (hidden) file on the datastore root called “.asyncUnmapFile”. This file could be taking up terabytes of space, and you aren’t able to delete this file.

asyncUnmapFile

Typically the asyncUnmapFile is used by the UNMAP feature on ESXi hosts to deal with unmapping and unallocating storage blocks on the SAN. When you run a manual UNMAP, this file should be created and should appear to using “0” (no) space (even if it is). When an UNMAP completes, this file should disappear and be automatically removed by the function. If an UNMAP is interupted, this file will not be deleted, allowing you to restart the process and upon a full successful completion, it should then be deleted.

The Problem

Some time ago, I had an issue when performing a manual UNMAP, where the ESXi host became unresponsive (due to memory issues). The command appeared to be completed, however I believe it caused potential issues or corruption on the iSCSI datastore. In subsequent runs, the UNMAP appeared to be functioning and working, however I didn’t realize that the asyncUnmapFile had grown to around 1.5TB.

This was noticed during a SAN storage audit, where we saw that the virtual pool on the SAN was using up way more storage than it should be on the datastore.

When we identified the file was this large and causing issues, we attempted to perform 2 UNMAPs (different reclaim sizes) to see if it would be automatically cleared afterwards. It had no effect and the file was unchanged.

We also tried to modify the permissions on the file, however when trying to delete it, it would report that the file or folder was not found, or that it does not exist. This was concerning as we were worried about potential datastore corruption.

It was also noticed that in the hostd.log and vmkernel.log we saw some errors where the host believed that the blocks on the datastore had already been freed: “on volume labeled ‘iSCSIDatastore01’ already freed by another host: This may be a non-issue”

The Solution

Unfortunately with all the research we did, we couldn’t find a clear-cut solution. With worries that the datastore may be corrupted, we needed to do something.

A decision was ultimately made to Storage vMotion all the VMs (Virtual Machines) to another datastore on a separate storage pool, delete the now empty LUN, and recreate it from scratch. After this, we used Storage vMotion again to move the VMs back.

Instantly I noticed that the VMs on that datastore were running faster (it’s only been 12 hours, so I’ll be adding an update in a few days to confirm). We no longer have the file on any of our datastores.

If anyone has further insight in to this issue, please leave a comment!

Procedure for restoring a VM from SAN Storage LUN snapshot

 ESXi, iSCSI, MSA, VMware, vSphere  No Responses »
Aug 272018
 
Right side of MSA 2040

So, what happens in a worst-case scenario where your backup system fails, you don’t have any VM snapshots, and the last thing standing in the way of complete data loss is your SAN storage systems LUN snapshots?

Well, first you fire whoever purchased and implemented the backup system, then secondly you need to start restoring the VM (or VMs) from your SAN LUN snapshots.

While I’ve never had to do this in the past (all the disaster recovery solutions I’ve designed and sold have been tested and function), I’ve always been curious what the process is and would be like. Today I decided to try it out and develop a procedure for restoring a VM from SAN Storage LUN snapshot.

For this test I pretended a VM was corrupt on my VMware vSphere cluster and then restored it to a previous state from a LUN snapshot on my HPE MSA 2040 (identical for the HPE MSA 2050, and MSA 2052) Dual Controller SAN.

To accomplish the restore, we’ll need to create a host mapping on the SAN for the LUN snapshot to a new LUN number available to the hosts. We then need to add and mount the VMFS volume (residing on the snapshot) to the host(s) while assigning it a new signature and then vMotion the VM from the snapshot’s VMFS to original datastore.

Important Notes (Read first):

  • When mounting a VMFS volume from a SAN snapshot, you MUST RE-SIGNATURE THE SNAPSHOT VMFS volume. Not doing so can cause problems.
  • The snapshot cannot be mapped as read only, VMFS volumes must be marked as writable in order to be mounted on ESXi hosts.
  • You must follow the proper procedure to gracefully dismount and detach the VMFS volume and storage device before removing the snapshot’s host mapping on the SAN.
  • We use Storage vMotion to perform a high-speed move and recovery of the VM. If you’re not licensed for Storage vMotion, you can use the datastore file browser and copy/move from the snapshot VMFS volume to live production VMFS volume, however this may be slower.
  • During this entire process you do not touch, modify, or change any settings on your existing active production LUNs (or LUN numbers).
  • Restoring a VM from a SAN LUN snapshot will restore a crash consistent copy of the VM. The VM when recovered will believe a system crash occurred and power was lost. This is NOT a graceful application consistent backup and restore.
  • Please read your SAN documentation for the procedure to access SAN snapshots, and create host mappings. With the MSA 2040 I can do this live during production, however your SAN may be different and your hosts may need to be powered off and disconnected while SAN configuration changes are made.
  • Pro tip: You can also power on and initialize the VM from the snapshot before initiating the storage vMotion. This will allow you to get production services back online while you’re moving the VM from the snapshot to production VMFS volumes.
  • I’m not responsible if you damage, corrupt, or cause any damage or issues to your environment if you follow these procedures.

We are assuming that you have already either deleted the damaged VM, or removed it from your inventory and renamed the VMs folder on the live VMFS datastore to change the name (example, renaming the folder from “SRV01” to “SRV01.bad”. If you renamed the damaged VM, make sure you have enough space for the new restored VM as well.

Procedure:

Mount the VMFS volume on the LUN snapshot to the ESXi host(s)
  1. Identify the VM you want to recover, write it down.
  2. Identify the datastore that the VM resides on, write it down.
  3. Identify the SAN and identify the LUN number that the VMFS datastore resides on, write it down.
  4. Identify the LUN Snapshot unique name/id/number and write it down, confirm the timestamp to make sure it will contain a valid recovery point.
  5. Log on to the SAN and create a host mapping to present the snapshot (you recorded above) to the hosts using a new and unused LUN number.
  6. Log on to your ESXi host and navigate to configuration, then storage adapters.
  7. Select the iSCSI initator and click the “Rescan Storage Adapters” button to rescan all iSCSI LUNs.

    VMware ESXi Host Rescan Storage Adapter

    VMware ESXi Host Rescan Storage Adapter

  8. Ensure both check boxes are checked and hit “Ok”, wait for the scan to complete (as shown in the “Recent Tasks” window.

    VMware ESXi Host Rescan Storage Adapter Window for VMFS Volume and Devices

    VMware ESXi Host Rescan Storage Adapter Window for VMFS Volume and Devices

  9. Now navigate to the “Datastores” tab under configuration, and click on the “Create a new Datastore” button as shown below.

    VMware ESXi Host Add Datastore Window

    VMware ESXi Host Add Datastore Window

  10. Continue with “VMFS” selected and select continue.
  11. In the next window, you’ll see your existing datastores, as well as your new datastore (from the snapshot). You can leave the “Datastore name” as is since this value will be ignored. In this window you’re going to select the new VMFS datastore from the snapshot. Make sure you confirm this by looking at the LUN number, as well as the value under “SnapshotVolume”. It is critical that you select the snapshot in this window (it should be the new LUN number you added above).
  12. Select next and continue.
  13. On the next window “Mount Option”, you need to change the radio button to and select “Assign a new signature”. This is critical! This will assign a new signature to differentiate it from your existing real production datastore so that the ESXi hosts don’t confuse it.
  14. Continue with the wizard and complete the mount process. At this point ESXi will resignture the VMFS volume and rename it to “snap-OriginalVolumeNameHere”.
  15. You can now browse the VMFS datastore residing on the LUN snapshot and do anything you’d normally be able to do with a normal datastore.
Copy/Move/vMotion the VM from the snapshot VMFS volume to your production VMFS volume

Note: The next steps are only if you are licensed for storage vMotion. If you aren’t you’ll need to use the copy or move function in the file browsing area to copy or move the VMs to your live production VMFS datastores:

  1. Now we’ll go to the vCenter/ESXi host storage area in the web client, and using the “Files” tab, we’ll browse the snapshots VMFS datastore that we just mounted.
  2. Locate the folder for the VM(s) you want to recover, open the folder, right click on the vmx file for the VM and select “Register VM”. Repeat this for any of the VMs you want to recover from the snapshot. Complete the wizard for each VM you register and add it to a host.
  3. Go back to you “Hosts and VMs” view, you’ll now see the VMs are added.
  4. Select and right click on the VM you want to move from the snapshot datastore to your production live datastore, and select “Migrate”.
  5. In the vMotion migrate wizard, select “Change Storage only”.
  6. Continue to the wizard, and storage vMotion the VM from the snapshot VMFS to your production VMFS volume. Wait for the vMotion to complete.
  7. After the storage vMotion is complete, boot the VM and confirm everything is functioning.
Gracefully unmount, detach, and remove the snapshot VMFS from the ESXi host, and then remove the host mapping from the SAN
  1. On each of your ESXi hosts that have access to the SAN, go to the “Datastores” section under the ESXi hosts configuration, right click on the snapshot VMFS datastore, and select “Unmount”. You’ll need to repeat this on each ESXi host that may have automounted the snapshot’s VMFS volume.
  2. On each of your ESXi hosts that have access to the SAN, go to the “Storage Devices” section under the ESXi hosts configuration and identify (by LUN number) the “disk” that is the snapshot LUN. Select and highlight the snapshot LUN disk, select “All Actions” and select “Detach”. Repeat this on each host.
  3. Double check and confirm that the snapshot VMFS datastore (and disk object) have been unmounted and detached from each ESXi host.
  4. You can now log in to your SAN and remove the host mapping for the snapshot-to-LUN. We will not longer present the snapshot LUN to any of the hosts.
  5. Back to the ESXi hosts, navigate to “Storage Adapters”, select the “iSCSI Initiator Adapter”, and click the “Rescan Storage Adapters”. Repeat this for each ESXi host.

    VMware ESXi Host Rescan Storage Adapter

    VMware ESXi Host Rescan Storage Adapter

  6. You’re done!
 Older Entries  Newer Entries