SOS runtime suspend/resume for multiple sessions accumulating into the same DBs
Closed this issue · 1 comments
cdwdirect commented
NOTE: Make sure to support differing allocation sizes w/each invocation.
cdwdirect commented
This works actually, but GUIDs get re-used and will mess up queries. Closing this, as a reformed GUID algorithm is described in a different issue.
what needs to happen is that IF the daemon is connecting to an existing database, during init we should query that database for the largest self-generated GUID, add 1, and start our range there. (at least for runs of the same size.) Either that, or all GUIDs' MSB's should be the job number, epoch, ...something to prevent collision in cases that daemon count differs (for aggregators).