RFR: 1939: Make work item scheduling more resilient against failures

[Erik Joelsson]

When the BotRunner calls getPeriodicItems, any failure will terminate the whole loop. This means that one misconfigured repository, or problem with the forge, can prevent all other repos from being processed. I think it would be a good idea to keep trying each bot instance in the list even if some of them fail.

This patch moves the catch clauses inside the loop. I'm also moving some logging and metrics gathering into a finally block.

-------------

Commit messages:
 - MACH5-1939

Changes: https://git.openjdk.org/skara/pull/1529/files
 Webrev: https://webrevs.openjdk.org/?repo=skara&pr=1529&range=00
  Issue: https://bugs.openjdk.org/browse/SKARA-1939
  Stats: 15 lines in 1 file changed: 8 ins; 7 del; 0 mod
  Patch: https://git.openjdk.org/skara/pull/1529.diff
  Fetch: git fetch https://git.openjdk.org/skara.git pull/1529/head:pull/1529

PR: https://git.openjdk.org/skara/pull/1529