iSCSI one session requested but one already present

W

WacArts

New Member
Proxmox Subscriber
Nov 26, 2024
6
0
1
Hi,

I have been having network issues for a while. They seem to be intermittent so cannot really put my finger on what is going wrong.
I did a complete shutdown and cold start from the network switches up to the nodes.

Background - running PVE 8.4 on 2 Dell servers and an MD storage array attached to each server using multipath.

This is working fine on one but the other is showing
iscsiadm: default: 1 session requested, but 1 already present. with pvesm status
This is then causing the array to error out with connection timed out or communication failure as it seems to be using the incorrect IP address to connect. It seems sometimes it uses a valid IP as two of the four addresses are correct but this is random.
All VMs are visible and accessible and I can ping all from each other and from outside PVE

iscsiadm -m node
172.16.0.11:3260,1 iqn..6782bcb0005fa0f10000000063daecea
172.16.0.12:3260,1 iqn..6782bcb0005fa0f10000000063daecea
172.16.0.13:3260,2 iqn..6782bcb0005fa0f10000000063daecea
172.16.0.14:3260,2 iqn..6782bcb0005fa0f10000000063daecea
172.16.1.11:3260,1 iqn..6782bcb0005fa0f10000000063daecea
172.16.1.12:3260,2 iqn..6782bcb0005fa0f10000000063daecea
172.16.2.11:3260,1 iqn..6782bcb0005fa0f10000000063daecea
172.16.2.12:3260,2 iqn..6782bcb0005fa0f10000000063daecea
172.16.0.21:3260,1 iqn..6f01faf000d706730000000063dafad8
172.16.0.22:3260,1 iqn..6f01faf000d706730000000063dafad8
172.16.0.23:3260,2 iqn..6f01faf000d706730000000063dafad8
172.16.0.24:3260,2 iqn..6f01faf000d706730000000063dafad8
172.16.1.21:3260,1 iqn..6f01faf000d706730000000063dafad8
172.16.1.22:3260,2 iqn..6f01faf000d706730000000063dafad8
172.16.2.21:3260,1 iqn..6f01faf000d706730000000063dafad8
172.16.2.22:3260,2 iqn..6f01faf000d706730000000063dafad8

As you can see, I am having duplicate links for the addresses and I cannot remove them. I have looked at the iSCSI controllers and they seem to be reporting correctly inasmuch that I cannot see any difference between the servers, the one that is working correctly and the one that isn't. The lines in bold are the correct addresses.

iscsiadm -m session -P1 is returning the same results on both servers


Any help/idea is gratefully appreciated
 
Update

I have tried iscsadm -m discoverydb --type sendtargets --portal <ip : port> --discover and the original IP addresses are found, the ones I am having issues with are not located and rtfm if the addresses are not located by the discover, they cannot be logged out from

Hence my dilemma
 
Last edited:
You must log in or register to reply here.
Top Bottom