@Danp Thanks for your response.

I already tried those steps but the problem is that I am getting authorisation error.

[17:17 xcp-ng-2 ~]# iscsiadm -m session
tcp: [10] 10.204.228.100:3260,1026 iqn.1992-08.com.netapp:sn.302cea4af64811ec84d9d039ea3ae4de:vs.5 (non-flash)
tcp: [7] 10.204.228.101:3260,1027 iqn.1992-08.com.netapp:sn.302cea4af64811ec84d9d039ea3ae4de:vs.5 (non-flash)
tcp: [8] 10.204.228.103:3260,1029 iqn.1992-08.com.netapp:sn.302cea4af64811ec84d9d039ea3ae4de:vs.5 (non-flash)
tcp: [9] 10.204.228.102:3260,1028 iqn.1992-08.com.netapp:sn.302cea4af64811ec84d9d039ea3ae4de:vs.5 (non-flash)

Not sure why its giving authetication error

[17:18 xcp-ng-2 ~]# iscsiadm -m discovery -t sendtargets -p 10.204.228.100
iscsiadm: Login failed to authenticate with target
iscsiadm: discovery login to 10.204.228.100 rejected: initiator failed authorization
iscsiadm: Could not perform SendTargets discovery: iSCSI login failed due to authorization failure
[17:18 xcp-ng-2 ~]#

If iscsi daemon has logged out then how I am able to access the VMs?
If iscsi daemon has logged out then how I am able to migrate VM from one xcp-ng node to another?

For migration and accessing VM, xcp-ng should be able to communicate SR, right?

However, I tried restart iscsi service in one of the xcp-ng node on which 2 VMs were there. One on local storage of the xcp-ng and another on the SR.
When I restarted the iscsid service and logged out and logged in to the storage, login failed giving authorisation error (not sure why it gave error, strange) and then both the VMs alongwith the xcp-ng node rebooted automatically
Once xcp-ng node came up, iscsi daemon automatically logged in to the storage.

I am not sure what to do in this case. Your support is much appreciated.

Hi All,

I am facing issue while taking backup using XOA.
" SR_BACKEND_FAILURE_202(, General backend error [opterr=rc: 21, stdout: , stderr: iscsiadm: No records found"

I am not even able to delete old snapshots from those VMs whose backups are failing. All these backups are on SR.

Need your support in getting this issue fixed.

Observation:

1. iscsi sessions are logged in from all the xcp-ng (hypervisor) nodes.
2. I am able to migrate VMs from one XCP-ng to another.
3. VM backup is happening only for those VMs which are residing on local HDD of the server and not on any SR.

Action taken:
I have restarted toolstack on all the nodes, still issue persists[2025-03-23T02_30_00.009Z - backup NG.json](Invalid file type. Allowed types are: .png, .jpg, .bmp, .txt, .diff, .patch, .jpeg) .

Log:
2025-03-23T02_30_00.009Z - backup NG.txt

@olivierlambert said in XO Lite in xcp-ng 8.2:

That's the other way around: it's a work and in progress, and -for example- actions on VMs are coming next week. We wanted to release very soon so people can start to play even it's a minimal version. Also, all that work will be re-used (partly) for XO 6.

Regarding master/slave: that's not a question of client (XCP-ng Center, XO Lite or XO): only the master can answer to a client. If your master is down, you have to elect a new one or use HA so it's done automatically. That's a design thing in XCP-ng/XenServer (all slaves are only in read only for the API database).

All the rest is a work in progress, expect new updates faster and faster, since the "longest" part is to design each component.

Are actions on VM delayed as I am unable to see any change yet

@olivierlambert Ok thank you
Will wait till next week to test the new rollout.

@olivierlambert said in XO Lite in xcp-ng 8.2:

That's exactly the point of XO Lite indeed If you have one pool with 4x hosts, the master will display all hosts in the pool.

If you have 4x pools with 1x host, you'll need to browser each host to get the info. As soon as you have more than 1x pool, XO is the tool to use.

Sorry to jump in here.
Hi Olivier, came across this XO Lite thread yesterday and today I was testing it in my setup.

First of all I want to thank you for this. Its really nice to see this tool, I can't express my happiness to you I was looking for some alternative of XCP-ng center as I am unable to use it due to security policies in one of my clients environment & my company is still thinking whether to buy the XOA license or not. As of now for every small tasks I need to visit DC which is a pain

This tool can be a lifesaver for me till I get XOA (if I get ) but when I logged in to the XO Lite, I can see all the XCP-ng servers in the pool and can also see all the VM's but unable to perform any task except working on VMs console. Hence below are my queries:-

1. Have you guys stopped working on this (as almost all the tabs are showing coming soon) or you are working on it but it will take some time to have a fully functional XO Lite? Any expected date for completion?
2. Similar to XCP-ng center XO lite also connects to master, so once master is down we cannot connect to the servers. Am I right?
3. In VM section, I can see many tabs but none of them are working, are they live or still work in progress.

Sorry for being too greedy