The nucleus services on 3.6.x are deployed as a standalone container on either the central node or a on a dedicated node.
The logs are persisted to disk and its location is deployment specific. They are usually under /opt or /domino. Also, the logs files are set to rotate every hr or so (again deployment specific), meaning it will be best to grab them immediately after the problem occurs.
Steps to grab the logs from disk, create a compressed tar ball and share with support :-
A small preview of tar and gzip : In Unix, files are stored in blocks of predefined size (1024, 4096). Each non-empty file, no matter how small it is, takes up one block (except for a completely empty file). So, a big file is generally better because it consumes all of the blocks completely wasting no space. This is what tar does, it takes a lot of small files and adds them into one big file but it does this by adding garbage chars at the end of the blocks to make them even sized hence we usually compress these archives to get rid of the repeated chars saving us at least 50% of the space.
- ssh into appropriate node from your bastion or directly if allowed
- cd to the logs location (mentioned above)
- next we archive the files and compress them :- tar -clf - . | gzip --best > ../"archive-name.tar.gz"
- next we copy the compressed archive to the bastion host :- (from the bastion) scp ubuntu@ip:path/to/file path/to/local (make sure you have apt permissions to the remote file)
- use pscp to copy the archive from bastion to local workstation (or you could use the mail builtin from dispatcher node, you could also do a curl to your smtp server and generate an email using that )
- send the file across to the support via email or by uploading it to a Domino deployment accessible by both the parties