Forum
Welcome, Guest
Username: Password: Remember me
This is the optional category header for the Suggestion Box.

TOPIC: XCP-ng support

XCP-ng support 3 months 1 week ago #1596

Hello any tests with xcp-ng and ha-lizard nosan???
The administrator has disabled public write access.

XCP-ng support 3 months 5 days ago #1597

We have not tested against xcp-ng. I would expect it to fully work as there are no changes to ng other than removing some license restrictions that I am aware of.
The administrator has disabled public write access.

XCP-ng support 2 months 1 week ago #1619

I've tested with the latest XCP-NG (7.4.1), works like a charm, there is no difference at all except the removing of licenced limitations from XenServer. The system worked even when i've upgraded one of my XenServer to XCP-ng (with reinstall Ha-Lizard components, ceartinly), while the other node runs Citrix XenServer :-)
The administrator has disabled public write access.
The following user(s) said Thank You: Nicolas, Diogo Serra

XCP-ng support 2 months 1 week ago #1620

Peter Timar wrote:
I've tested with the latest XCP-NG (7.4.1), works like a charm, there is no difference at all except the removing of licenced limitations from XenServer. The system worked even when i've upgraded one of my XenServer to XCP-ng (with reinstall Ha-Lizard components, ceartinly), while the other node runs Citrix XenServer :-)

I have one pool with two hosts xenserver 7.2.0 and ha-lizard nosan can you guide me how do you do the update?
Last time i check i already have the latest ha-lizard version

Thanks
The administrator has disabled public write access.

XCP-ng support 1 month 4 weeks ago #1629

This will be really helpfull, also have 2 xen6.5 with halizard drbd and have to move on to iscsi with XCP-ng
The administrator has disabled public write access.

XCP-ng support 1 month 4 weeks ago #1630

The method is similar (or equal), when you replace a dead node from the cluster. In a nutshell:

1. designate the pool master role to your "survivor" node
2. disable ha (if you don't disable ha, the dead node will be automatically kikcked, but just in case...). The iscsi-ha does not need to set manual.
3. halt old node and kick it from the pool
3. upgrade your old node (when you upgrade, your network setup will not be altered)
4. after upgrade join to HA-Pool
5. install NOSAN package (watch for DRBD master/slave IP-s!) - if your setup changed (i.e. the partitions has changed, then change iscsi-ha to manual, and correct your drbd setup, but after a simple upgrade there will be no changes).
6. when NOSAN install completed, no need to attach SR (SR is already in the pool).
7. install ha-lizard, as usual
8. enable ha and wait for DRBD sync complete

If you done well, you have an upgraded pool member with zero downtime. If you want to upgrade the another node, repeat the whole procedure again.

Please do this tasks first on a test environment, I wrote this as I remember...

I've not tested on 6.5, I don't know the same method is works. There will be more problem (different SR, different DRBD versions), but I believe this is not impossible. Interesting, maybe I make a try... :-)
The administrator has disabled public write access.
Time to create page: 0.112 seconds