I just got bit by this today. I tracked it down to the default solr.xml file
in ./server/solr/solr.xml with the following:
<shardHandlerFactory name="shardHandlerFactory"
class="HttpShardHandlerFactory">
<int name="socketTimeout">${socketTimeout:600000}</int>
<int name="connTimeout">${connTimeout:60000}</int>
</shardHandlerFactory>
I’m on Solr 5.3.1 now, and I wasn’t having this problem with 4.10.3, and sure
enough, 4.10.3 has the values at 0 (i.e. no socket timeout)
-David
> On Jan 24, 2016, at 3:18 PM, Shawn Heisey <apa...@elyograg.org> wrote:
>
> On 1/23/2016 9:24 PM, adfel70 wrote:
>> 1. I am getting the "read time out" from the Solr Server.
>> Not from my client, but from the server client when it tries to reach other
>> instances while committing.
>>
>> 2. I reduced the filter cache autowarmCount to 512, and seems to fix the
>> problem. It now takes only several seconds to commit!
>
> Do you have any configuration for ShardHandler in your solrconfig.xml?
>
> https://wiki.apache.org/solr/SolrConfigXml#Configuration_of_Shard_Handlers_for_Distributed_searches
>
> This is where the client built into Solr can be configured with a socket
> timeout.
>
> Regarding your cache configuration, even an autowarmCount of 512 is
> quite high. I have configured a value of *four* for my filterCache,
> because anything higher resulted in unacceptable commit times. You may
> need to experiment with your configuration for best results.
>
> Thanks,
> Shawn
>
