We're currently using MySQL Aurora DB, and despite not observing any explicit bad connection closures, we're encountering an issue related to reading communication packets, which in turn is causing a spike in the aborted clients metric on our RDS MySQL Aurora database.
[NOTE] Got an error reading communication packets (sql_connect.cc:828)
I'm also interested in understanding the potential risks associated with a spike in the aborted clients metric.
We're leveraging Hikari for managing our thread pool and timeouts, and I'm seeking recommendations on potential adjustments within Hikari configurations to address these issues without directly altering MySQL parameters.
Given that this problem is occurring in production database environment, I'm wondering if there's a way to simulate similar conditions locally to experiment with various Hikari configurations. Additionally, I'm curious if there are any known issues with the MySQL version we're using.
I appreciate any insights or suggestions on how to approach and resolve these challenges. Thanks in advance for your assistance.
AWS OFFICIELA mis à jour il y a 7 mois