There are some monitoring tools which take a snapshot of the process table every few minutes. If you happen to run one of those, and your process happens to be running when the snapshots is taken you can catch it.

If this is a postgresql server and you have logging turned up with pids being logged, and the pin in question was a database process, you can probably find it in the database logs.

Sometimes after upgrades, even minor ones, I find it useful to run analyze on all of the tables. I usually do analyze verbose; so I can see which tables are getting analyzed. This will assess every table so the query planner can make better decisions about how to resolve queries. If the query planner is making bad decisions I/O and CPU will be high and query performance will be poor.