You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-user@lucene.apache.org by Daniel Brügge <da...@googlemail.com> on 2012/05/16 20:29:58 UTC
CloudSolrServer not working with standalone Zookeeper
Hi,
I am just playing around with SolrCloud and have read in articles like
http://www.lucidimagination.com/blog/2012/03/05/scaling-solr-indexing-with-solrcloud-hadoop-and-behemoth/that
it
is sufficient to create the connection to the Zookeeper instance and not to
the Solr instance.
When I try to connect to my standalone Zookeeper instance (not started
with a Solr instance and "-DzkRun") I am getting this error:
Caused by: java.util.concurrent.TimeoutException: Could not connect to
> ZooKeeper
I am also getting this error when I try to connect directly to one of the
Solr instances.
My code looks like this:
solr = new CloudSolrServer("myzkhost:2181");
((CloudSolrServer) solr).setDefaultCollection("collection1");
I am working with the latest Solr trunk version (
https://builds.apache.org/view/S-Z/view/Solr/job/Solr-trunk/1855/)
Do I need to start the zookeeper in Solr to keep this working?
Thanks & regards
Daniel
Re: CloudSolrServer not working with standalone Zookeeper
Posted by Daniel Brügge <da...@googlemail.com>.
Ok, it seems that a maven dependency to zookeeper version 3.3 broke this.
Now it connects to the zk instance.
Thanks.
On Mon, May 21, 2012 at 5:31 PM, Daniel Brügge <
daniel.bruegge@googlemail.com> wrote:
> Thanks for your feedback. I don't know.
>
> I've tried just now with the newest trunk version and the embedded ZK on
> port 9983.
>
> In the logs of the zk-solr it shows:
>
> *INFO: Accepted socket connection from /XXX.XXX.XXX.XXX:1055*
> *May 21, 2012 3:27:34 PM org.apache.zookeeper.server.NIOServerCnxn doIO*
> *WARNING: EndOfStreamException: Unable to read additional data from
> client sessionid 0x0, likely client has closed socket*
> *May 21, 2012 3:27:34 PM org.apache.zookeeper.server.NIOServerCnxn
> closeSock*
> *INFO: Closed socket connection for client /XXX.XXX.XXX.XXX:1055 (no
> session established for client)*
>
>
> So it can definitely connects to the port in my opinion, but it closes the
> connection after the defined timeout (here 10000ms)
>
> *Caused by: java.util.concurrent.TimeoutException: Could not connect to
> ZooKeeper MYZKHOST.:9983 within 10000 m*
>
> Hmm. I also thought that this trivial setup should work. Will check again.
>
> Daniel
>
> On Fri, May 18, 2012 at 4:23 PM, Mark Miller <ma...@gmail.com>wrote:
>
>> Seems something is stopping the connection from occurring? Tests are
>> constantly running and doing this using an embedded zk server - and I know
>> more than a few people using an external zk setup. I'd have to guess
>> something in your env or URL is causing this?
>>
>>
>> On May 16, 2012, at 3:11 PM, Daniel Brügge wrote:
>>
>> > OK, it's also not working with an internal started Zookeeper.
>> >
>> > On Wed, May 16, 2012 at 8:29 PM, Daniel Brügge <
>> > daniel.bruegge@googlemail.com> wrote:
>> >
>> >> Hi,
>> >>
>> >> I am just playing around with SolrCloud and have read in articles like
>> >>
>> >>
>> http://www.lucidimagination.com/blog/2012/03/05/scaling-solr-indexing-with-solrcloud-hadoop-and-behemoth/thatit
>> >> is sufficient to create the connection to the Zookeeper instance and
>> not
>> >> to the Solr instance.
>> >> When I try to connect to my standalone Zookeeper instance (not started
>> >> with a Solr instance and "-DzkRun") I am getting this error:
>> >>
>> >> Caused by: java.util.concurrent.TimeoutException: Could not connect to
>> >>> ZooKeeper
>> >>
>> >>
>> >> I am also getting this error when I try to connect directly to one of
>> the
>> >> Solr instances.
>> >>
>> >> My code looks like this:
>> >>
>> >> solr = new CloudSolrServer("myzkhost:2181");
>> >> ((CloudSolrServer)
>> solr).setDefaultCollection("collection1");
>> >>
>> >> I am working with the latest Solr trunk version (
>> >> https://builds.apache.org/view/S-Z/view/Solr/job/Solr-trunk/1855/)
>> >>
>> >> Do I need to start the zookeeper in Solr to keep this working?
>> >>
>> >> Thanks & regards
>> >>
>> >> Daniel
>> >>
>>
>> - Mark Miller
>> lucidimagination.com
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
Re: CloudSolrServer not working with standalone Zookeeper
Posted by Daniel Brügge <da...@googlemail.com>.
Thanks for your feedback. I don't know.
I've tried just now with the newest trunk version and the embedded ZK on
port 9983.
In the logs of the zk-solr it shows:
*INFO: Accepted socket connection from /XXX.XXX.XXX.XXX:1055*
*May 21, 2012 3:27:34 PM org.apache.zookeeper.server.NIOServerCnxn doIO*
*WARNING: EndOfStreamException: Unable to read additional data from client
sessionid 0x0, likely client has closed socket*
*May 21, 2012 3:27:34 PM org.apache.zookeeper.server.NIOServerCnxn closeSock
*
*INFO: Closed socket connection for client /XXX.XXX.XXX.XXX:1055 (no
session established for client)*
So it can definitely connects to the port in my opinion, but it closes the
connection after the defined timeout (here 10000ms)
*Caused by: java.util.concurrent.TimeoutException: Could not connect to
ZooKeeper MYZKHOST.:9983 within 10000 m*
Hmm. I also thought that this trivial setup should work. Will check again.
Daniel
On Fri, May 18, 2012 at 4:23 PM, Mark Miller <ma...@gmail.com> wrote:
> Seems something is stopping the connection from occurring? Tests are
> constantly running and doing this using an embedded zk server - and I know
> more than a few people using an external zk setup. I'd have to guess
> something in your env or URL is causing this?
>
>
> On May 16, 2012, at 3:11 PM, Daniel Brügge wrote:
>
> > OK, it's also not working with an internal started Zookeeper.
> >
> > On Wed, May 16, 2012 at 8:29 PM, Daniel Brügge <
> > daniel.bruegge@googlemail.com> wrote:
> >
> >> Hi,
> >>
> >> I am just playing around with SolrCloud and have read in articles like
> >>
> >>
> http://www.lucidimagination.com/blog/2012/03/05/scaling-solr-indexing-with-solrcloud-hadoop-and-behemoth/thatit
> >> is sufficient to create the connection to the Zookeeper instance and not
> >> to the Solr instance.
> >> When I try to connect to my standalone Zookeeper instance (not started
> >> with a Solr instance and "-DzkRun") I am getting this error:
> >>
> >> Caused by: java.util.concurrent.TimeoutException: Could not connect to
> >>> ZooKeeper
> >>
> >>
> >> I am also getting this error when I try to connect directly to one of
> the
> >> Solr instances.
> >>
> >> My code looks like this:
> >>
> >> solr = new CloudSolrServer("myzkhost:2181");
> >> ((CloudSolrServer) solr).setDefaultCollection("collection1");
> >>
> >> I am working with the latest Solr trunk version (
> >> https://builds.apache.org/view/S-Z/view/Solr/job/Solr-trunk/1855/)
> >>
> >> Do I need to start the zookeeper in Solr to keep this working?
> >>
> >> Thanks & regards
> >>
> >> Daniel
> >>
>
> - Mark Miller
> lucidimagination.com
>
>
>
>
>
>
>
>
>
>
>
>
Re: CloudSolrServer not working with standalone Zookeeper
Posted by Mark Miller <ma...@gmail.com>.
Seems something is stopping the connection from occurring? Tests are constantly running and doing this using an embedded zk server - and I know more than a few people using an external zk setup. I'd have to guess something in your env or URL is causing this?
On May 16, 2012, at 3:11 PM, Daniel Brügge wrote:
> OK, it's also not working with an internal started Zookeeper.
>
> On Wed, May 16, 2012 at 8:29 PM, Daniel Brügge <
> daniel.bruegge@googlemail.com> wrote:
>
>> Hi,
>>
>> I am just playing around with SolrCloud and have read in articles like
>>
>> http://www.lucidimagination.com/blog/2012/03/05/scaling-solr-indexing-with-solrcloud-hadoop-and-behemoth/that it
>> is sufficient to create the connection to the Zookeeper instance and not
>> to the Solr instance.
>> When I try to connect to my standalone Zookeeper instance (not started
>> with a Solr instance and "-DzkRun") I am getting this error:
>>
>> Caused by: java.util.concurrent.TimeoutException: Could not connect to
>>> ZooKeeper
>>
>>
>> I am also getting this error when I try to connect directly to one of the
>> Solr instances.
>>
>> My code looks like this:
>>
>> solr = new CloudSolrServer("myzkhost:2181");
>> ((CloudSolrServer) solr).setDefaultCollection("collection1");
>>
>> I am working with the latest Solr trunk version (
>> https://builds.apache.org/view/S-Z/view/Solr/job/Solr-trunk/1855/)
>>
>> Do I need to start the zookeeper in Solr to keep this working?
>>
>> Thanks & regards
>>
>> Daniel
>>
- Mark Miller
lucidimagination.com
Re: CloudSolrServer not working with standalone Zookeeper
Posted by Daniel Brügge <da...@googlemail.com>.
OK, it's also not working with an internal started Zookeeper.
On Wed, May 16, 2012 at 8:29 PM, Daniel Brügge <
daniel.bruegge@googlemail.com> wrote:
> Hi,
>
> I am just playing around with SolrCloud and have read in articles like
>
> http://www.lucidimagination.com/blog/2012/03/05/scaling-solr-indexing-with-solrcloud-hadoop-and-behemoth/that it
> is sufficient to create the connection to the Zookeeper instance and not
> to the Solr instance.
> When I try to connect to my standalone Zookeeper instance (not started
> with a Solr instance and "-DzkRun") I am getting this error:
>
> Caused by: java.util.concurrent.TimeoutException: Could not connect to
>> ZooKeeper
>
>
> I am also getting this error when I try to connect directly to one of the
> Solr instances.
>
> My code looks like this:
>
> solr = new CloudSolrServer("myzkhost:2181");
> ((CloudSolrServer) solr).setDefaultCollection("collection1");
>
> I am working with the latest Solr trunk version (
> https://builds.apache.org/view/S-Z/view/Solr/job/Solr-trunk/1855/)
>
> Do I need to start the zookeeper in Solr to keep this working?
>
> Thanks & regards
>
> Daniel
>