lucene-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sachin Kale <sachinpk...@gmail.com>
Subject Re: Master-Slave setup using SolrCloud
Date Wed, 01 Oct 2014 17:01:53 GMT
Actually, the image was captured from SOLR dashboard having following lines:

*Error    SolrCore                  java.lang.NullPointerException*
*Error    SolrDispatchFilter    java.lang.NullPointerException*

We are now on SolrCloud setup completely.
Memory allocated to each SOLR instance is 7GB. I guess it is more than
sufficient for 1 GB index, right?
Indexes are stored as normal, local filesystem.
We are using latest SOLR version, 4.10.0
I am using three caches:
    Query Cache: Size 4096, autoWarmCount 2048
    Filter cache: size 8192, autoWarmCount 4096
    Document cache: size 4096

I think I am setting wrong values from commit maxTime. Just found out
following document:
http://lucidworks.com/blog/understanding-transaction-logs-softcommit-and-commit-in-sorlcloud/

According to this document, if our requirement is Index-Heavy, Query-Heavy,
then autoSoftCommit should be as long as high (not sure what should be its
value in my case) and hardCommit to 15 seconds (I am using 15 minutes).
Should I change my hard commit interval to 15 minutes?

On Wed, Oct 1, 2014 at 9:58 PM, Mark Bennett <mark.bennett@lucidworks.com>
wrote:

> Hello Sachin,
>
> FYI, at least in my email client, I don't see the embedded graphic (shows
> a broken image link).  Not sure if that's on my end, or maybe the mailing
> list server.
>
> To your question, some clarification please:
>
> * You mentioned moving to SolrCloud, but are you still currently running
> v3 master/slave, or is this now a SolrCloud setup?
>
> * Is it possible you're running low on memory?  Could you maybe share the
> startup switches (memory, GC, etc)
>
> * A long shot, but I have to ask, where are the indexes being stored?  is
> this a "normal", local filesystem, or shard storage, or some type of more
> exotic storage?  (HDFS, etc)
>
> * Exact version of Solr?
>
> * The NPE's (Null Pointer Exceptions), are those on the master or
> replicas?  Have you tried temporarily running with fewer replaces, perhaps
> just 1, for temp debugging purposes?
>
> Mark
>
> --
> Mark Bennett / LucidWorks: Search & Big Data / mark.bennett@lucidworks.com
> <mailto:mark.bennett@lucidworks.com>
> Office: 408-898-4201 / Telecommute: 408-733-0387 / Cell: 408-829-6513
>
> On Oct 1, 2014, at 8:49 AM, Sachin Kale <sachinpkale@gmail.com<mailto:
> sachinpkale@gmail.com>> wrote:
>
> Hello,
>
> We are trying to move our traditional master-slave Solr configuration to
> SolrCloud. As our index size is very small (around 1 GB), we are having
> only one shard.
> So basically, we are having same master-slave configuration with one
> leader and 6 replicas.
> We are experimenting with maxTime of both AutoCommit and AutoSoftCommit.
> Currently, autoCommit maxTime is 15 minutes and autoSoftCommit is 1 minute
> (Let me know if these values does not make sense).
>
> Caches are set such that warmup time is at most 20 seconds.
>
> We are having continuous indexing requests mostly for updating the
> existing documents. Few requests are for deleting/adding the documents.
>
> The problem we are facing is that we are getting very frequent
> NullPointerExceptions.
> We get continuous 200-300 such exceptions within a period of 30 seconds
> and for next few minutes, it works fine.
>
>
> ‚Äč
> I am not sure what would be causing it. My guess, whenever, it is trying
> to replay tlog, we are getting these exceptions. Is anything wrong in my
> configuration?
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message