Working on upgrade planning for move from vSphere 5.5 to 6.7. Using iSCSI storage (on a Pure Storage array.) Discovered that ESXi 6.5/7 will not recognize a volume increase until the host is restarted. Not an ideal workaround. I don't see this behavior on ESXi 6.0, or 5.5, both which recognize the increase following a rescan of the storage adapter. Has anyone else run into this? Found a better workaround?
Steps to reproduce:
Storage
→ Devices
and the datastore (if applicable) under Storage
→ Datastore
Storage
→ Devices
→ Rescan
Workaround:
Storage
→ Devices
→ observe "Capacity" column now reflects size increaseStorage
→ Devices
→ observe "Capacity" column now reflects size increaseI've seen this too, and to me it looks like a nasty bug, because it's present from all clients (Embedded Host Client, vSphere Web Client, and the HTML5 client).
What worked for me was to run vmkfstools -V locally on all hosts.
André
- Increase the size in your Storage array first. (Pure storage?)
- Right click the datastore in the vSphere client and click Increase capacity...(3-step wizard, done)
Sadly, this doesn't work either. The wizard doesn't show any available space, so I'm back to restarting the host so it will see that indeed there IS more space on that iSCSI volume.
Just for comparison's sake, I did try your suggestion on ESXi 6.0 and 5.5, and it works: the wizard sees the additional space and let's me expand the datastore. Looks like my issue is isolated to ESXi 6.5 and 6.7.
[Edited for spelling]
I've seen this too, and to me it looks like a nasty bug, because it's present from all clients (Embedded Host Client, vSphere Web Client, and the HTML5 client).
What worked for me was to run vmkfstools -V locally on all hosts.
André
Thanks so much for this suggestion, André, it worked for me but I've got no idea why it should since the -v
option only enables verbose output of the vmkfstools
command (documentation here.) Definitely smells like a bug.
Confirmed your suggestion did work for me, as seen in the following output. Notice the volume is first shown as 500GB, but after using the -V
option it suddenly shows up at 1TB, which is the correct size. Your workaround solves my problem. I really appreciate the help.
[root@KNHESXI29T:~] vmkfstools -Ph /vmfs/volumes/ALPHA/
VMFS-5.60 (Raw Major Version: 14) file system spanning 1 partitions.
File system label (if any): ALPHA
Mode: public Capacity 499.8 GB, 266.6 GB available, file block size 1 MB, max supported file size 62.9 TB
UUID: 5cf85ed7-65267ec6-b1b9-000c29f8cdb5
Partitions spanned (on "lvm"): naa.624a9370967d557ce5c84d1c000113e9:1
Is Native Snapshot Capable: YES
[root@KNHESXI29T:~] vmkfstools -V
[root@KNHESXI29T:~] vmkfstools -Ph /vmfs/volumes/ALPHA/
VMFS-5.60 (Raw Major Version: 14) file system spanning 1 partitions.
File system label (if any): ALPHA
Mode: public Capacity 1023.8 GB, 499.7 GB available, file block size 1 MB, max supported file size 62.9 TB
UUID: 5cf85ed7-65267ec6-b1b9-000c29f8cdb5
Partitions spanned (on "lvm"): naa.624a9370967d557ce5c84d1c000113e9:1
Is Native Snapshot Capable: YES
[Edited for formatting.]
You are right, the "-v" (lower case) just produces a verbose output, but the "-V" (upper case) performs a refresh of the VMFS file system.
André
Yes, it was a BUG.
However latest version of ESXi 6.5 host has all the fixes