[SOLVED] Physical volume on other hosts

[adamb]

I have some iscsi based storage with LVM ontop. Introduced a new lun this morning to all 7 nodes.

On one of the hosts I created a physical volume and volume group on mpath6.

Iscsi looks good and so does multipath on all the hosts. Each host has the following.

mpath6 (2b53371228193c3eb6c9ce9003d3584d2) dm-686 Nimble,Server
size=127T features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
|- 28:0:0:0 sdi 8:128 active ready running
`- 29:0:0:0 sdj 8:144 active ready running


On the other hosts I can see the physical volume and volume group with pvscan and vgscan, but I can't get them to actually show up.

root@ccscloud4:~# pvscan
PV /dev/mapper/mpath3 VG Cloud-Alletra lvm2 [<127.00 TiB / <10.54 TiB free]
PV /dev/mapper/mpath5 VG Cloud-Transient-Alletra lvm2 [<100.00 TiB / <53.46 TiB free]
PV /dev/mapper/mpath6 VG Cloud-Udarchive1 lvm2 [<127.00 TiB / <127.00 TiB free]
PV /dev/mapper/mpath4 VG Cloud2-Alletra lvm2 [<127.00 TiB / 87.03 TiB free]
PV /dev/sda3 VG pve lvm2 [<279.24 GiB / <15.81 GiB free]

/dev/mapper/mpath6 is what I just added.

On the host I created the pv and vg.

root@ccscloud1:~# pvdisplay /dev/mapper/mpath6
--- Physical volume ---
PV Name /dev/mapper/mpath6
VG Name Cloud-Udarchive1
PV Size 127.00 TiB / not usable 4.00 MiB
Allocatable yes
PE Size 4.00 MiB
Total PE 33292287
Free PE 33292287
Allocated PE 0
PV UUID jOE3SR-Y2no-ZDgI-vISO-4G9o-gl7q-gHekXS

On any of the other hosts in the cluster.

root@ccscloud4:~# pvdisplay /dev/mapper/mpath6
Failed to find physical volume "/dev/mapper/mpath6".

Not sure what I could be missing, we have added luns in the past without hitting any of this.

 

[bbgeek17]

Sometimes things dont work out due to timing and you need to give it an extra kick. A reboot will most likely solve the issue.
You can also try to experiment with "pvscan" options, ie -t -d -v -a (see man pvscan)
Check you journalctl as well.

---

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

 

[adamb]

Sometimes things dont work out due to timing and you need to give it an extra kick. A reboot will most likely solve the issue.
You can also try to experiment with "pvscan" options, ie -t -d -v -a (see man pvscan)
Check you journalctl as well.

---

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

That was enough to get me on the right track.

This did the job.

pvscan --cache

 

[rgeoffroy]

thank you very much, the pvscan --cache command really worked.