12.2.01 Adding Node Fails w/ Verifying '/tmp/' ...FAILED

From: David Barbour <david.barbour1_at_gmail.com>
Date: Tue, 26 May 2020 09:25:21 -0500
Message-ID: <CAFH+ifeoNFXzLK1iiFruUJ_pGbpYbzRntC3h5nGtBB0iGLxv9g_at_mail.gmail.com>

Good Morning,

I am trying to add a node to our 2-node RAC cluster running RHEL 6.8 and Oracle The new node is running RHEL 7.6. This should not be an issue as I have an internal RAC cluster where one node is on 7.6 and the other on 6.8 (pending OS upgrade).

When I run cluvfy, ( cluvfy stage -post hwos -n 685921-db5,685925-db6,1103477-db7 -verbose ) I get the following error: Failures were encountered during execution of CVU verification request
"stage -post hwos".

Verifying '/tmp/' ...FAILED
1103477-db7: Version of exectask could not be retrieved from node

I've checked the Oracle notes and adjusted the MTU size on the new node, created a dummy file on one of the current RAC nodes and successfully scp'd it over to /tmp as the grid user on the new node.

Googled the daylights out of this as well. No luck.

I'm thinking I'm missing a package or something, but the lib* rpms all seem to be at or above the requisite levels per Oracle documentation.

Can anyone suggest what might be a problem here?

Received on Tue May 26 2020 - 16:25:21 CEST

Original text of this message