2

I have a RDS instance running in AWS. By default

max_connections=45

so I used mysqlslap to see how it works. I ran:

mysqlslap --concurrency=20 --host=deleteme.ckn5fsf3nwqp.ca-central-1.rds.amazonaws.com --user=deleteme --password --auto-generate-sql

and it worked nicely. Then I modified the --concurrency to 45, 50, 100, 150, 200 and it worked anyway, I was expecting a 'Too many connections' error, but I only see that error from --concurrency to over 300.

So my 1st question is, if my max_connections is set to 45, and using a concurrency of ~200 connections, how doesn't mysqlslap give me error?

I also tried this experiment running mysql in an EC2 instance. I tried --concurrency with 500 connections, and got the 'Too many connections error', but that error is not logged in /var/log/mysql. Is it a default configuration that mysql doesn't log those errors?

and my last question is again regarding RDS. I was hoping to see in cloudwatch, a spike on DB connections, but even using mysqlslap at concurrency=45, I only see 'DB Connection Count' metric to 1.

Maybe mysqlslap concurrency does not create new connections?

  • While mysqlslap was running more than 45 connections, what was in SHOW PROCESSLIST? What did these status values say: Threads_running, Max_used_connections, Connection_errors_max_connections, Aborted_clients, Aborted_connects? – Rick James Jun 7 '18 at 2:32
  • What kinds of queries were run? Look at the Com% STATUS values before and after -- to see how many were actually run. – Rick James Jun 7 '18 at 2:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.