Sharing another pxgrid issue in case others encounter the same thing. The cluster recently upgraded to 1.3.3.5, and we were unable to exchange CTS/GBAC data from ISE (main driver) into DNAC even though the pxgrid subscribers in ISE were both 'ON' and DNAC showed no ill signs of issues. However, working with TAC we identified that the GBAC configuration within DNAC reset itself to a sync status of Day 0. This essentially means you are required to force(start) GBAC migration to DNAC, and then if you wish back to ISE. Also, note that we were advised that a previous workaround of modifying the AAA server password to force a re-sync with ISE is no longer advised and will trigger a GBAC reset back to 'Day 0'. Prior to starting GBAC migration you can do the following to tail the logs:
In DNAC UI go to System Settings->Debugging Logs->Enable the following:
Then from CLI run the following: magctl service logs -rf aca- | tee ACA-service
After starting GBAC migration the DB acaworkflowtable changes from this:
(excluding additional DB text as you should probably work with TAC)
-[ RECORD 1 ]+---------------
importstatus | REQUIRED_DAY_0
selectedmode | UNKNOWN
To this (DNAC Driver):
-[ RECORD 1 ]+----------
importstatus | COMPLETED
selectedmode | ACA
Finally to this if you flip back to ISE:
-[ RECORD 1 ]+----------
importstatus | COMPLETED
selectedmode | ISE
Once this is done pxgrid data sharing (ISE/DNAC Sync Status) is working as expected. Should you see the 'start migration banner' appear in Policy->GBAC again, and are unable to share data via pxgrid even though things seem good then you are likely encountering the same issue. Lastly, apparently this is scheduled to be fixed in 1.3.3.7. Here is the bug relating to this issue: https://bst.cloudapps.cisco.com/bugsearch/bug/CSCvu52300/?rfs=iqvred
HTH!
