Hi Rowan,
you understood very well and I really appreciate your answer.
I've tested callback repl_queue_info without parameter name and it provides some data - replication is running so everything seems to be synced. I'll retest it with NAS deactivated on HA HUB.
But to be honest I'm really interested in the fact we could turn on nis_bridge on HA HUB after failover. We don't use UMP_failover script from communities, however we've written our script some time ago that, in fact, does the same - change UIM paths from primary to HA HUB in UMP configuration and restart UMP. As UNIX/Linux admins and being not so experienced in NAS/LUA scripting we've written the script in bash and start it regularly from cron.
Do I understand it well that after UMP failover, i.e. UIM paths are changed, we could enable nis_bridge on HA HUB? In all documents it is recommended to disable it on HA HUB. Moreover, we were worried about alarm duplication in NIS database after failback from HA back to primary HUB. When NAS starts to replicate alarms from HA HUB back to primary HUB and primary HUB will write them once more to NIS database. Or am I wrong?
Enabling nis_bridge would probably help us with two issues we're facing when UIM is running on HA HUB, they are
-1) "green" UMPs, they show no alarm, probably due to the fact that trellis depends on nis_bridge
-2) some NAS/LUA scripts which read data from NAS_* tables in NIS database don't work properly
Kind regards,
Marek Lukes