Zbx 6LTS to 6.4 Docker Zwazing host item interval rates. /StartTimers
Hi.
We've just migrated from 6LTS running on mysql natively on Ubuntu to a 6.4 Docker based solution, also using mysql, and have transitioned the mysql database and most of our server.conf parameters (poller numbers etc) over, and everything appears to work except three items on one host. The one parameter we've not copied over is StartTimers, which was 5 based on 'try stuff', but reverted to the default of 1 because it seems, from the manual, to only a maintenance tasks, not active poll tasks. Host Memory, processor speed and disk access rates have all massively improved too in the new host.
Those three items are a userparameter that runs a custom dos batch file on the host, which typically takes less than 1 second to complete.
The items were queried with a simple update interval of 20 seconds, and worked reliably prior to the migration. Following the migration, a random one of these will often sit unsampled for 2 minutes while the others updates about every 20s.
Server and agent Debug at 3 doesn't show these items failing : they just seem entirely unqueried.. I've tried setting updates to s0-59/20 amongst other values and it's had no affect.
While I'm going to boost StartTimers back to 5 in the morning, just to rule it out, and increasing debug to 4, but is there anything else I should be looking at?
(will likely go zbx 7.0 Lts via docker if this all works out)
Ta!
We've just migrated from 6LTS running on mysql natively on Ubuntu to a 6.4 Docker based solution, also using mysql, and have transitioned the mysql database and most of our server.conf parameters (poller numbers etc) over, and everything appears to work except three items on one host. The one parameter we've not copied over is StartTimers, which was 5 based on 'try stuff', but reverted to the default of 1 because it seems, from the manual, to only a maintenance tasks, not active poll tasks. Host Memory, processor speed and disk access rates have all massively improved too in the new host.
Those three items are a userparameter that runs a custom dos batch file on the host, which typically takes less than 1 second to complete.
The items were queried with a simple update interval of 20 seconds, and worked reliably prior to the migration. Following the migration, a random one of these will often sit unsampled for 2 minutes while the others updates about every 20s.
Server and agent Debug at 3 doesn't show these items failing : they just seem entirely unqueried.. I've tried setting updates to s0-59/20 amongst other values and it's had no affect.
While I'm going to boost StartTimers back to 5 in the morning, just to rule it out, and increasing debug to 4, but is there anything else I should be looking at?
(will likely go zbx 7.0 Lts via docker if this all works out)
Ta!
Comment