Database load and performance

{question}

Why is it that when the database has an issue, it is because of load or multiple queries or long running queries?

{question}

{answer}

You can collect all of the following information and send it to NuoDB Support Team for analysis:

Check whether there are specific hours of the day that you have more load, at this range of time, we can we check periodically the number of connections, by:
                       SELECT count(*) from system.connections;

We can check which queries ran longer than 60 seconds since the last restart:
                       SELECT SQLSTRING from System.querystats Where RUNTIME>60000000;

Collect the output from the Linux “top” command.

You should create a diagnose, by:
/opt/nuodb/bin/nuodbmgr --broker localhost --password <password> --command "diagnose domain outdir <output_dir> fullcore true"

Monitor memory usage using the “monitor … “ command

  • nuodbmgr --command “monitor …”

The NuoDB metric “Memory” shows the amount of memory taken by a nuodb engine (TE or SM) process:

  • nuodbmgr --command “monitor domain” | grep Memory

         See: http://doc.nuodb.com/Latest/Default.htm#Metrics.htm

The NuoDB metric “HeapActive” shows the size of the internal cache heap:

  • nuodbmgr --command “monitor domain” | grep HeapActive

{answer}

Have more questions? Submit a request

Comments