MariaDBMon, ReadWriteSplit + max_slave_replication_lag, and MySQL RDS

[Wagner Bianchi]

Hello Folks,

Do you know about any possible issue when running MaxScale for a simple master/slave when this pair of backends are MySQL RDS? 

Also, I see the below for my current replica on the MaxScale log file:

2021-08-31 21:51:58.508   warning: (2709) (rwsplit-service) (set_rlag_state): Replication lag of 'dbreplica' is -1 seconds, which is above the configured limit 30s. 'dbreplica' is excluded from query routing.

As this is a test environment, the replica isn't lagging, and I wonder if this is related to the GTID diffs between MariaDB and MySQL. 

My basic config for the ReadWriteSplit Router is below:

[maxscale]

threads=auto

log_augmentation=1

ms_timestamp=1

admin_host=0.0.0.0

admin_port=8989

admin_secure_gui=false

writeq_high_water=16M

writeq_low_water=128K

[rwsplit-service]

type=service

router=readwritesplit

user=maxusr

password=FFC7...

max_sescmd_history=300

prune_sescmd_history=true

master_reconnection=true

master_failure_mode=fail_on_write

max_slave_replication_lag=30s

use_sql_variables_in=master

Any hints? Thanks!

[Markus Mäkelä]

Hi,

I think we recently saw that same error message and came to the conclusion that it was a sort of a false alarm: at the same time a message about the server losing the Slave status should've been logged. Even if the message isn't logged, the value -1 means that a valid value for the replication lag wasn't extracted which hints at some problem in either authentication between MaxScale and the database or in the replication itself.

Markus

--
You received this message because you are subscribed to the Google Groups "MaxScale" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxscale+u...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/maxscale/5f489f42-198e-40b2-b410-deee46fdbfbfn%40googlegroups.com.

--

Markus Mäkelä, Senior Software Engineer
MariaDB Corporation

[Wagner Bianchi]

It does make sense, Markus. After performing the changes proposed on this thread, it worked fine.

root@maxscale01: ~ # maxctrl show services | grep max_slave_replication

│                     │     "max_slave_replication_lag": 30000,                     │

I'll hammer the master and check if after lagging that much, the replica will be removed from the rotation.

Thanks; you rock, dude.