16Gb RAM / 8 cores : best config?


(Yann Barraud) #1

Hi

What would be the best cubic for ES running on a dedicated machine with 16
Gb of RAM & 8 cores ?

Thanks,
Yann

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/7c91a40b-c99e-4894-961d-f99f01803a34%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Jörg Prante) #2

No easy answer, it depends on your OS and your requirements - indexing
load, searching load, analytics ... you should start with ES default
settings and heap size increased to 50% of RAM, that is 8G

Jörg

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CAKdsXoFi_LSygR1WSJhetfo97WjoE_niZ1T2r%2BpP%2BvHCJMfE3w%40mail.gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Yann Barraud) #3

Hi,

Thanks for the answer. I have an index with 2,5M docs, and 8,5 Go. I'm
doing fuzzylikethisfield queries, and custom score queries.

OS is ubuntu, virtualized.

And the results are really slow. As if it was running an a single core...

Cordialement,
Yann Barraud

2013/12/31 joergprante@gmail.com joergprante@gmail.com

No easy answer, it depends on your OS and your requirements - indexing
load, searching load, analytics ... you should start with ES default
settings and heap size increased to 50% of RAM, that is 8G

Jörg

--
You received this message because you are subscribed to a topic in the
Google Groups "elasticsearch" group.
To unsubscribe from this topic, visit
https://groups.google.com/d/topic/elasticsearch/r5wx8u2prns/unsubscribe.
To unsubscribe from this group and all its topics, send an email to
elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/elasticsearch/CAKdsXoFi_LSygR1WSJhetfo97WjoE_niZ1T2r%2BpP%2BvHCJMfE3w%40mail.gmail.com
.

For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CA%2BhvuXd820HakDfKT%3DkWZf3Kzp_a5q6seyZJ4-fVBUXyGra6HA%40mail.gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Jörg Prante) #4

Do not use VM if you want maximum single node performance.

And, fuzzylikethis is slow, yes. Have you tried the parameters
minSimilarity, prefixLength, ignoreTF, maxnumterms?

Or use another query type which is faster?

Jörg

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CAKdsXoErV1%2Bb%3DQrx-ZQodzDJAGMjC%3DAxE0f6iR0rBtEUrcqmPg%40mail.gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Yann Barraud) #5

Hi,

I will take a look at this. I can't have nothing but a VM ATM. And I think
fuzzylikethis is the good one for me (running this bvefore rescoring with
my dedup plugin). But I did not have a look at those params. Thanks for
pointing me, I'll let you know...

Cordialement,
Yann Barraud

2014/1/1 joergprante@gmail.com joergprante@gmail.com

Do not use VM if you want maximum single node performance.

And, fuzzylikethis is slow, yes. Have you tried the parameters
minSimilarity, prefixLength, ignoreTF, maxnumterms?

Or use another query type which is faster?

Jörg

--
You received this message because you are subscribed to a topic in the
Google Groups "elasticsearch" group.
To unsubscribe from this topic, visit
https://groups.google.com/d/topic/elasticsearch/r5wx8u2prns/unsubscribe.
To unsubscribe from this group and all its topics, send an email to
elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/elasticsearch/CAKdsXoErV1%2Bb%3DQrx-ZQodzDJAGMjC%3DAxE0f6iR0rBtEUrcqmPg%40mail.gmail.com
.

For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CA%2BhvuXfzftwVBDYJtc%2BjAW8Rjggre8UF_3Fy%3DkNd-_ewQHL84Q%40mail.gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.


(system) #6