Raven 3.5 Voron index reduce time

10 views
Skip to first unread message

Kyle Levien

unread,
May 21, 2018, 2:35:45 PM5/21/18
to RavenDB - 2nd generation document database
We have a couple multi-map reduce index that appears to be taking a very long time per batch to reduce. These indexes are fairly sizable, but I'm seeing what looks like a lot of 'idle' time in the indexing stats, where I see total durations of over a minute, but individual ms timings of just a couple hundred (screencap attached). Any idea what might be causing this? Does the server implement thread waiting if it's being hammered too much?


Thanks,
Kyle

Kyle Levien

unread,
May 21, 2018, 2:41:50 PM5/21/18
to RavenDB - 2nd generation document database
As a follow up, I tried stopping and restarting the server and see index times starts quick but fairly quickly ramp back up to over a minute per batch. Here's a capture of one of the faster earlier batches:

Oren Eini (Ayende Rahien)

unread,
May 21, 2018, 2:49:49 PM5/21/18
to ravendb
Are you running multiple indexes at the same time? 

Hibernating Rhinos Ltd  

Oren Eini l CEO l Mobile: + 972-52-548-6969

Office: +972-4-622-7811 l Fax: +972-153-4-622-7811

 


--
You received this message because you are subscribed to the Google Groups "RavenDB - 2nd generation document database" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ravendb+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Kyle Levien

unread,
May 21, 2018, 3:05:40 PM5/21/18
to RavenDB - 2nd generation document database
When I first saw the issue, yes it was just after initializing all our indexes, but it was down to just two when I stopped and restarted the server and got the repro.
To unsubscribe from this group and stop receiving emails from it, send an email to ravendb+u...@googlegroups.com.

Kyle Levien

unread,
May 21, 2018, 3:07:11 PM5/21/18
to RavenDB - 2nd generation document database

It is down to just the one index and this is what I see after a restart (increasing index size, seemingly exponential to batch size)

Oren Eini (Ayende Rahien)

unread,
May 21, 2018, 3:08:57 PM5/21/18
to ravendb
Can you try this on 4.0 as well?

Hibernating Rhinos Ltd  

Oren Eini l CEO l Mobile: + 972-52-548-6969

Office: +972-4-622-7811 l Fax: +972-153-4-622-7811

 


On Mon, May 21, 2018 at 10:07 PM, Kyle Levien <teh...@gmail.com> wrote:

It is down to just the one index and this is what I see after a restart (increasing index size, seemingly exponential to batch size)

--
You received this message because you are subscribed to the Google Groups "RavenDB - 2nd generation document database" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ravendb+unsubscribe@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages