This is what I'm doing in the Java API for search;
And even if I put a 2 sec timeout, I have often requests that take more
than 5 and even 10 seconds that does not time out. But sometime, I have
some query that take under 2 sec and says timed out.
On the documentation it says;
timeout A search timeout, bounding the search request to be executed within
the specified time value and bail with the hits accumulated up to that
point when expired. Defaults to no timeout.
With that I'm expecting every query to be under (or roughly) around 2
second, but not 5 or 10 seconds.
Am I doing something wrong, or is it really and ES issue ?
I have seen this issue pop up on the mailing list every now and then. The
timeout value is used only by the TimeLimitingCollector, so it only limits
the time used by Lucene in collecting documents. The parsing/gather/reduce
phases all lie outside of this time interval. However, I would find it
strange that something like the reduce phase would take so long.
This is what I'm doing in the Java API for search; Timeout · GitHub
And even if I put a 2 sec timeout, I have often requests that take more
than 5 and even 10 seconds that does not time out. But sometime, I have
some query that take under 2 sec and says timed out.
On the documentation it says;
timeout A search timeout, bounding the search request to be executed
within the specified time value and bail with the hits accumulated up to
that point when expired. Defaults to no timeout.
With that I'm expecting every query to be under (or roughly) around 2
second, but not 5 or 10 seconds.
Am I doing something wrong, or is it really and ES issue ?
Possibly because I use a lot of facetting on the same query ? Facetting and
sorting is done on the reduce phase if I'm not wrong ?
On Thursday, January 3, 2013 12:28:40 PM UTC-5, Ivan Brusic wrote:
I have seen this issue pop up on the mailing list every now and then. The
timeout value is used only by the TimeLimitingCollector, so it only limits
the time used by Lucene in collecting documents. The parsing/gather/reduce
phases all lie outside of this time interval. However, I would find it
strange that something like the reduce phase would take so long.
For the record, I have the same issue.
--
Ivan
On Thu, Jan 3, 2013 at 8:47 AM, Jérôme Gagnon <jerome....@gmail.com<javascript:>
wrote:
This is what I'm doing in the Java API for search; Timeout · GitHub
And even if I put a 2 sec timeout, I have often requests that take more
than 5 and even 10 seconds that does not time out. But sometime, I have
some query that take under 2 sec and says timed out.
On the documentation it says;
timeout A search timeout, bounding the search request to be executed
within the specified time value and bail with the hits accumulated up to
that point when expired. Defaults to no timeout.
With that I'm expecting every query to be under (or roughly) around 2
second, but not 5 or 10 seconds.
Am I doing something wrong, or is it really and ES issue ?
Facets are implemented with FacetCollectors, which means they are computed
at search time. The TimeLimitingCollector is a simple wrapper around the
collectors that do the real work. The same goes for sorting, although there
has to be logic during the reduce phase that resorts and aggregates facets.
Possibly because I use a lot of facetting on the same query ? Facetting
and sorting is done on the reduce phase if I'm not wrong ?
On Thursday, January 3, 2013 12:28:40 PM UTC-5, Ivan Brusic wrote:
I have seen this issue pop up on the mailing list every now and then.
The timeout value is used only by the TimeLimitingCollector, so it only
limits the time used by Lucene in collecting documents. The
parsing/gather/reduce phases all lie outside of this time interval.
However, I would find it strange that something like the reduce phase would
take so long.
And even if I put a 2 sec timeout, I have often requests that take more
than 5 and even 10 seconds that does not time out. But sometime, I have
some query that take under 2 sec and says timed out.
On the documentation it says;
timeout A search timeout, bounding the search request to be executed
within the specified time value and bail with the hits accumulated up to
that point when expired. Defaults to no timeout.
With that I'm expecting every query to be under (or roughly) around 2
second, but not 5 or 10 seconds.
Am I doing something wrong, or is it really and ES issue ?
On Thursday, January 3, 2013 11:47:08 AM UTC-5, Jérôme Gagnon wrote:
This is what I'm doing in the Java API for search; Timeout · GitHub
And even if I put a 2 sec timeout, I have often requests that take more
than 5 and even 10 seconds that does not time out. But sometime, I have
some query that take under 2 sec and says timed out.
On the documentation it says;
timeout A search timeout, bounding the search request to be executed
within the specified time value and bail with the hits accumulated up to
that point when expired. Defaults to no timeout.
With that I'm expecting every query to be under (or roughly) around 2
second, but not 5 or 10 seconds.
Am I doing something wrong, or is it really and ES issue ?
On Thursday, January 3, 2013 11:47:08 AM UTC-5, Jérôme Gagnon wrote:
This is what I'm doing in the Java API for search; Timeout · GitHub
And even if I put a 2 sec timeout, I have often requests that take more
than 5 and even 10 seconds that does not time out. But sometime, I have
some query that take under 2 sec and says timed out.
On the documentation it says;
timeout A search timeout, bounding the search request to be executed
within the specified time value and bail with the hits accumulated up to
that point when expired. Defaults to no timeout.
With that I'm expecting every query to be under (or roughly) around 2
second, but not 5 or 10 seconds.
Am I doing something wrong, or is it really and ES issue ?
On Thursday, January 3, 2013 11:47:08 AM UTC-5, Jérôme Gagnon wrote:
This is what I'm doing in the Java API for search; Timeout · GitHub
And even if I put a 2 sec timeout, I have often requests that take more
than 5 and even 10 seconds that does not time out. But sometime, I have
some query that take under 2 sec and says timed out.
On the documentation it says;
timeout A search timeout, bounding the search request to be executed
within the specified time value and bail with the hits accumulated up to
that point when expired. Defaults to no timeout.
With that I'm expecting every query to be under (or roughly) around 2
second, but not 5 or 10 seconds.
Am I doing something wrong, or is it really and ES issue ?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.