rudybrewster
Contributor
Contributor

HA fails to run on cluster after upgrading 1 host to 4.1

Jump to solution

Hi  I had a 3 server cluster, all running ESX3.5 with DRS & HA enabled.  Last week I upgraded the VI server to vCenter 4.1, and all seemed ok. Tonight I tried to upgrade one host via the update manager, I remediated the server, it went through ok, but when the 4.1 server came back the cluster couldnt configure HA anymore. The error told me there were insufficient resources, but I don't think that's the case, the vCenter server has plenty of disk space, RAM etc.  Looking on the forums the most common issue was the length of the FQDN, but mine should be ok at 27 characters.   so as it stands I have no HA, but I was able to move a VM fro server 2 (3.5) to server 3 (3.5) any idea why that works?  Can anyone help?  Thanks

0 Kudos
1 Solution

Accepted Solutions
bilalhashmi
Expert
Expert

Great.. now your current failover capacity also seems to be in line with what is configured... awesome..:)

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal

View solution in original post

0 Kudos
21 Replies
bilalhashmi
Expert
Expert

Not sure if this is realted to you issue.. have a look

http://virtualcitizen.org/2011/01/05/upgrade-esxi-3-5-to-4-1-ha-configuration-fails-a-general-system...

Can you vMotion your VMs?

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal
0 Kudos
rudybrewster
Contributor
Contributor

It could be, I'll give that a try, thanks.  my cluster says HA is not working, but as mentioned I could vMotion a VM from host 2 (Still ESX3.5) to host 3 (still ESX3.5) I cannot move any VMs onto host one (newly upgraded to 4.1)

0 Kudos
bilalhashmi
Expert
Expert

What error o u get when u try to vMotion to the new host? Are all the portgroups and network settings in place for vMotion to work? shared storage etc etc?

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal
0 Kudos
DSTAVERT
Immortal
Immortal

Try moving the host out of the cluster and then move it back in.

-- David -- VMware Communities Moderator
rudybrewster
Contributor
Contributor

All of the shared storage seems to be there, all my datastores etc are on the summary page, all the networking seems to be there too (it should be as I upgraded via Update Manager)

Should I right click > Remove then right click the cluster name > Add host, and see if that works? Is that the best way to remove/readd?

error.JPG

error2.JPG

0 Kudos
bilalhashmi
Expert
Expert

hould I right click > Remove then right click the cluster name > Add host, and see if that works? Is that the best way to remove/readd?

Yes, your host will have to be in maintenance mode though.

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal
0 Kudos
bilalhashmi
Expert
Expert

BTW, what is your HA failover setting?

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal
0 Kudos
rudybrewster
Contributor
Contributor

Do you mean these settings?

setting.JPG

0 Kudos
DSTAVERT
Immortal
Immortal

Yes I would give it a try as you suggest.

-- David -- VMware Communities Moderator
0 Kudos
rudybrewster
Contributor
Contributor

I'll try removing - if I remove the host from the cluster, will I lose all of the settings I have configured? There's a lot of stuff set-up, so I'd like to avoid this if possible.

0 Kudos
a_p_
Leadership
Leadership

Did you verify the disk space on the ESX host itself yet? Please run df -h to make sure there is no full file system.

André

0 Kudos
bilalhashmi
Expert
Expert

Yes, so ur failover capacity is configured to be at 1. And admission control is enabled which will ensure no more slots are used than what will enable u to have HA when needed.. Your current failover capacity is 0 which means u can't afford to loose any more hosts...  Not sure if that is adding up in my head..

If you can't get it to work by readding it to the cluster, can u try disabling HA in the cluster and then renabling HA, assuming u dont have a lot of custom settings for HA in ur cluster?

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal
bilalhashmi
Expert
Expert

What are those settings that you are worried about.. just curious..

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal
0 Kudos
rudybrewster
Contributor
Contributor

Ok I've lost the ability to SSH to the host, so I can't run the disk check to see how much space is, I'll have to wait until tomorrow for that.

I'm disabling HA on the cluster and going to re enable and see if that works.

0 Kudos
rudybrewster
Contributor
Contributor

Ok I've reconfigured HA and it seems happy at the moment Smiley Happy

My settings look a little healthier:

HAnow.JPG

I guess I'll leave it overnight and  check it in the morning, if all is well I'll give host 2 a try...

Thanks everyone for all of your help.

0 Kudos
DSTAVERT
Immortal
Immortal

Glad things are working. Please do update the post with your results.

-- David -- VMware Communities Moderator
0 Kudos
rudybrewster
Contributor
Contributor

Will do, I have dished out all the helpful answers points I can, I'll give someone correct answer points tomorrow if all is well.

0 Kudos
DSTAVERT
Immortal
Immortal

That you award points or not isn't as important as someone with a similar issue can search and find the string of things that work.

-- David -- VMware Communities Moderator
0 Kudos
bilalhashmi
Expert
Expert

Great.. now your current failover capacity also seems to be in line with what is configured... awesome..:)

Follow me @ Cloud-Buddy.com

Blog: www.Cloud-Buddy.com | Follow me @hashmibilal

View solution in original post

0 Kudos