Intermittent Operational Error 2013 - Lost connection during query... many times a day and apart from that it behaves normally.
Infra. Overview:
We're using K8s cluster (created using kops) to deploy our applications but our primary DB server is not yet migrated to kubernetes which means it still resides in a separate EC2 instance outside kubernetes cluster.
We use Flask_SQLAlchemy to connect our flask app to MySQL.
I've created mysql service in my namespace with endpoint pointing to private IP of Mysql's EC2 instance.
As I know, similar questions related to OperationalError have been already asked and I've applied almost all of the fixes suggested.
sqlalchemy_pool_recycle = <way-less-than-servers-wait_timeout> // 900 (mysql's is 28800)sqlalchemy's pre_ping is set to truenf_conntrack_tcp_timeout_close_wait = 3600 (but pool will be recycled in 900)Please help me to understand it deeper. Suggestions welcomed.
its flask app config issue. you can update in app config below settings that should fix the issue.
SQLALCHEMY_POOL_SIZE = 50 # scale app for pool size # default 10
SQLALCHEMY_POOL_RECYCLE = 250 # pool recycle time in seconds Flask_sqlAlchemy has issue to close session , you need update in mysql config as well .
[mysql]
lock_wait_timeout=300It should less than SQLALCHEMY_POOL_RECYCLE as mentioned in flask documentation here :- https://flask-sqlalchemy.palletsprojects.com/en/2.x/config/#timeouts
"it is recommended that you set SQLALCHEMY_POOL_RECYCLE to a value less than your backend’s timeout".