Prevent 'Aborted connection' warnings from the DoctrineConnectionHealthCheck

[pablothedude]

Aborted connection warnings where spawned into the messages log on every call to the health endpoint. Many sleeping mysql tasks remained unclosed.

To resolve this issue, the DoctrineConnectionHealthCheck needed to be changed. The previously used execute statement turned out to cause the problem. Using the query method solved the problem.

[thijskh]

In my test, does not seem to have the desired effect.

[tvdijen]

Nope, indeed still loads of aborted db-connections from engine, middleware and gateway:

Sep 11 20:14:23 database-1 mysqld: 2018-09-11 20:14:23 1702 [Warning] Aborted connection 1702 to db: 'engineblock' user: 'engineblock' host: '10.2.16.3' (Got an error reading communication packets)
Sep 11 20:14:51 database-1 mysqld: 2018-09-11 20:14:51 1703 [Warning] Aborted connection 1703 to db: 'gateway' user: 'gateway' host: '10.2.16.3' (Got an error reading communication packets)
Sep 11 20:15:19 database-1 mysqld: 2018-09-11 20:15:19 1705 [Warning] Aborted connection 1705 to db: 'middleware' user: 'middleware' host: '10.2.16.2' (Got an error reading communication packets)

[pablothedude]

We've tested the issue on the test environment and could reproduce the problem there. The problem was that the wrong method was used and therefore the connections were not properly closed. This fix should now address this issue.

[thijskh]

Yes, works better now!

[MKodde]

@pablothedude, would you be so kind to install this change in:

• Stepup-*
• SP Dashboard
• User lifcecycle
• Engine (even tho this change was already fixed in the overloaded Checker)

See: https://www.pivotaltracker.com/story/show/158906723