Forum: Other

Post: org.voltcore.logging.VoltUtilLoggingLogger log INFO: 3617 estimated time update.?

org.voltcore.logging.VoltUtilLoggingLogger log INFO: 3617 estimated time update.?
matrix
Aug 12, 2014
Hi, everyone!

I'm using VoltDB for my course project. And I noticed that when VoltDB is in use, it sends info like:
Aug 12, 2014 6:43:05 PM org.voltcore.logging.VoltUtilLoggingLogger log INFO: 3617 estimated time update.

This is not an exception or an error, but VoltDB tends to be slower around the occurrence of this INFO. I don't know why. Could anyone explain this? Thanks!

===============my deployment.xml===================
<?xml version="1.0"?>
<deployment>
<cluster hostcount="1" sitesperhost="6" />
<commandlog enabled="false" />
</deployment>
nshi
Aug 12, 2014
Hi,

This message is printed when the time update thread has not run for more than 2 seconds. Normally, the thread is scheduled to run once every 5 millisecond.

Are you running VoltDB inside a VM on a laptop? The threads may be contending for resource. Other threads may also be delayed which may cause the database to slow down.
matrix
Aug 13, 2014
Hi, Thanks for reply!

I'm running VoltDB inside a VM.

How does this matter? Is there anything I can do to enable the time update thread?
nshi
Aug 13, 2014
The time update thread not being scheduled for such a long time is just a symptom of resource contention, it is not the cause of the slowdown. If the VM does not have enough resource, other threads that do real work in Volt may not be scheduled for a long period of time, which will slow down the system.

What is the hardware spec of the host system and how much resource did you allocate for the VM? What edition and version of Volt were you running?
matrix
Aug 14, 2014
The host system has two four-core CPUs and 64GB memory. The VM is allocated all CPU resource and 16GB memory. I'm running VoltDB ent edition, version 4.5.

I'm running a performance test using both VoltDB and MySQL. There are nearly 500 Voltdb tables which are accessed by around 100 threads concurrently. At the same time, 500 threads launch queries to MySQL continually.

I guess CPU contention is the cause of the slowdown, as too many threads are running concurrently, not to mention the host running whatever other programs.

Thank you, ning!