Client reached out with the following on how they should handle the following behavior:
Having an issue with “Source Directory” messages specifically with the “Group Filter Entry”. My observation is that when a TREP server goes down (F9: Server has gone down) I’m receiving Group-Id’s with Status (non-merge) that are not in my UPA Connection.
I can replicate the problem by subscribing to one instrument. When I receive the Refresh message the Group-Id is logged. I also log all entry points in UPA, namely Refresh (Solicted/UnSolicted), Source Directory and Status messages. I then use adsmon “Source Routes and Application Program” to bring down Server.
As an actual example using my attached logs;
- 2023-06-23T11:02:41.165709-0400
- Refresh w/ Group-Id 9 // GOOD GROUP
- 2023-06-23T11:02:41.165709-0400
- adsmon bring Server down
- Source Update Group-Id 2 w/ Status Suspect // BAD GROUP
- 2023-06-23T11:02:41.166545-0400
- Merge Group-Id 2 to Group-Id 0 // BAD GROUP
- 2023-06-23T11:02:41.167225-0400
- Source Update Group-Id 3 w/ Status Suspect
- 2023-06-23T11:02:41.167860-0400
- Merge Group-Id 3 -> Group-Id 0
- 2023-06-23T11:02:41.168479-0400
- Text -> F9: Server has gone down.
- 2023-06-23T11:02:41.169131-0400
- Merge Group-Id 9 -> Group-Id 0 // GOOD GROUP MERGE
All logs are attached to email. Can you please advise on the solution to this issue
- Merge Group-Id 9 -> Group-Id 0 // GOOD GROUP MERGE
You can use FeedDataServer.sort.txt (time sorted) to view Refresh and Source Update messages and sources.UPA_OPT_A-ZZZZZ.*.txt files for individual Source messages (Releated to
FeedDataServer.sort.txt)
sources.UPA_OPT_A-ZZZZZ.refresh.000000-105410.647166.txt
sources.UPA_OPT_A-ZZZZZ.update.000001-110241.165953.txt
sources.UPA_OPT_A-ZZZZZ.update.000002-110241.166642.txt