migration problem

[bradawk1]

We are trying to migrate all of our servers from an 8.5 applicance to another 8.5 appliance.  The servers are on the same subnet.  I put in the correct adminstrator account and password and I click on connect.  A few minutes later, it pops up "Unable to connect to source appliance."  From the settings, network section, I can ping the other appliance.  So, I know connectivity works.  Any other ideas?

[bradawk1 · Accepted solution]

Never mind.  I think the problem I could not log in was because of some corruption in the original appliance database.  I used the rest api and just force added all of the hardware to the new appliance.  I already copied over the server profile templates.  Next is the licenses. 

[ChrisLynch]

Glad you were able to make progress.  You can get the list of license keys and what "nodes" they are assigned to using GET /rest/lienses.  In the members collection, you will see an object per license.  Within that object, there is a nodes collection that lists the nodeName of server nodes associated with the key.

[bradawk1]

Hi Chris,

When I started to look at the licenses, I remembered that the licenses were tied to the keys in each iLO.  So, the licenses followed the iLOs.  So, there was nothing left to do on that front.  However, were still experiencing two problems.  One was that when we clicked on the management ip address, it failed to open the connection.  If I put the ip address into the address bar, it would open it with no problem (except I would have to manually log in).  I found that if I changed the :

<IPV6_PREFERRED_PROTOCOL VALUE="N"/>

using hponcfg, that behaviour stopped.

Now we just have to figure out why we keep getting :

HPE OneView did not receive and expected REST event notification from the server's iLO <ilo name>.

It is on most of the nearly 600 servers.

[ChrisLynch]

HPE OneView did not receive and expected REST event notification from the server's iLO <ilo name>.

This manifests itself in one of two ways:

• Older iLO firmware
• Workarounds within OV to overcome component Redfish issue wasn't working in certain cases

What OV and iLO firmware version are you running?

[bradawk1]

Hi Chris,

All of our iLOs are Gen10 or Gen10+ (a few 3par ssmc iLOs are gen 9) and have 2.96 installed.  OneView is 8.5.  We can update to 8.6 if you think that would help?

[ChrisLynch]

It could help, but don't update to 8.60 just yet.  We are in the final stages of re-releasing 8.60 due to an issue with OVRS generating messages that it was unable to connect to HPE.

[bradawk1]

We updated two of the XL225n servers to 2.99 iLO firmware and are downloading 8.60.01 now.  So far, the two servers with 2.99 aren't behaving any better than the rest with 2.96.  Maybe after we update to 8.60.01?

[bradawk1]

@ChrisLynch that did not make a difference.  A lot of the errors I saw in the ciDebug log from the support dump were in reference to SNMP settings like:

Unable to access the ILO from the HPE OneView Appliance
Error while verifying snmp Events unable to configure SNMP settings on the server iLO

Question is, how do I tell OneView to re-generate/re-deploy SNMP settings to all iLOs?

[bradawk1]

Also, last week we started sending the appliance audit logs to splunk.  Just checked this morning (the appliance had crashed) and saw many messages with contents:

Stopping appliance because of an unexpected error

or

Restarting appliance to recover from an unexpected problem.

[ChrisLynch]

Question is, how do I tell OneView to re-generate/re-deploy SNMP settings to all iLOs?

A server refresh with credentials API call (PUT /rest/server-hardware/{id}/refreshState) will reconfigure/fix management settings on the iLO.

Just checked this morning (the appliance had crashed) and saw many messages with contents:

I would suggest you work that specific issue with HPE Support via a support case.

[bradawk1]

Hi Chris,

I do have a case open.