Aborted Connections In Error Log

[HeebyGeeby]

I’m seeing an aborted connection error in my MySQL error log, once every 5 minutes, presumably every time MonYog connects. 

 

The MySQL server is new, on a new server, the config is fairly vanilla.

 

Both MonYog and MySQL appear to be running fine, other than the error log.

 

The wait_timeout is the default 28800. 

I’ve tried setting net_read_timeout and net_write_timeout to high values (500 secs), makes no difference. 

 

MonYog 6.5.2 windows.

MySQL is 5.6.29 on Centos 7.2 

 

An example MySql error log line is:

2016-03-31 17:07:00 24296 [Warning] Aborted connection 1872 to db: ‘unconnected’ user: ‘monyog.user’ host: ‘my.public.ip’ (Got timeout reading communication packets)

 

There seems to be a corresponding log line in MonYog:

[6.52] [2016-03-31 17:06:00] [Server: Azure DB01] srcymysql.cpp(705) ErrCode:2013 ErrMsg:Lost connection to MySQL server during query

 

The error seems to appear in the MonYog log first.

 

Could Centos be killing the thread?

[HeebyGeeby]

Solved my own problem.

 

 

My db server is hosted on Microsoft Azure.

 

 

I believe the cause is this:

 

http://serverfault.com/questions/661704/azure-closing-idle-network-connections

 

 

Azure has a 4 min timeout for load balancers, which apparently also affects networks that don’t use a load balancer. So it’s an Azure “feature”.

 

It’s possible to increase this timeout within Azure.

 

[Author]

Posts