Translating with DrWatson… this can take a few seconds the first time.
This is a genuine, complex translation. DrWatson protects commands, error codes, and log output while naturally translating the surrounding text. It’s translated once and saved.
DAN MUELLER — — source: IIUG Forums & Mailing Lists
IDS 12.10.FC9
AIX 7.2
Morn'n Folks,
I want to be able to assure that my HDR/RSS replication is healthy and in
sync. My idea was to create a table with one datetime year to second column,
insert 1st systables.sysdate into that table on the primary then query on the
HDR & RSS from the other 2 servers, collect the outputs and diff them. If they
all match - I figure I am cool.
My question is... has someone else already figured out something to check
replication is in sync? Obviously, I am trying to avoid doing this manually
and I want to cron it so it runs as much as I like and creates an incident for
me if there is any mismatch.
TIA,
Dan
Dan,
IMO not really much you can do other than rely on HDR/RSS faith along with
instance monitoring (onstat -g dri , DRLOSTANDFOUND and ALARMPROGRAM). If
really concerned consider periodic HDR refreshes if feasible.
CDR has cdr check, maybe a feature request for something similar for HDR/RSS
is in line ?
As for the table test, it only checking that that one particular table is in
sync and doesn't prove much other than that. For the test to be considered
significant you would have to perform "random sampling" from a population of
other such test tables and if reported to be in sync then you could conclude
from statistical inference that all tables are in sync.
Mark
↪ replying to DAN MUELLER
LUIS MARQUES — — source: IIUG Forums & Mailing Lists
You can look at the scheduler task "ifx_ha_monitor_log_replay_task" in the
sysadmin:ph_task . It calls the procedure "ifx_ha_monitor_log_replay" which
you can look at and maybe use it as a starting point for what you want?
How about comparing output of sysmaster:syslogs on both sides ?
normally, the current log (is_current=1) should have the same number on both
sides and the 'used' column should have approximate values.
If you don't have the same current log # on both sides or the 'used' is 'much
lower' on the secondary side, this probably means you have latency.
Never tried, but this should work
We use strictly necessary cookies to make this site work. With your
consent we’d also use optional cookies for analytics and marketing. You can accept all,
reject all, or choose. Read our Cookie Policy.