Architecture for scaling ES

Hi all,

I need to know how scale ES.
I have 15millions docs for one index, 5millions for another one, and I do a
lot of big faceting (terms stats, etc...).

How many servers I need? I know it's difficult to say something like that,
but right now, I have 2 servers, 4 core and 7.5 go RAM each one.
It's really slow.

Any idea?

Thanks

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

In fact, for users who knows AWS, I'm on EC2, with m1.large instances

On Wednesday, September 18, 2013 12:05:18 PM UTC+2, Fabien Guiraud wrote:

Hi all,

I need to know how scale ES.
I have 15millions docs for one index, 5millions for another one, and I do
a lot of big faceting (terms stats, etc...).

How many servers I need? I know it's difficult to say something like that,
but right now, I have 2 servers, 4 core and 7.5 go RAM each one.
It's really slow.

Any idea?

Thanks

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

Hey,

you havent said anything about your configuration or you definition of slow
and even your expectations. Maybe you can elaborate about all this bit
more...

Generally the best thing is to start with one server and see when it
degrades in performance (this might be indxing speed, query latency or
whatever metric you have defined for yourself) and then start slowly with
scaling out.

--Alex

On Wed, Sep 18, 2013 at 12:09 PM, Fabien Guiraud fabien1909@gmail.comwrote:

In fact, for users who knows AWS, I'm on EC2, with m1.large instances

On Wednesday, September 18, 2013 12:05:18 PM UTC+2, Fabien Guiraud wrote:

Hi all,

I need to know how scale ES.
I have 15millions docs for one index, 5millions for another one, and I do
a lot of big faceting (terms stats, etc...).

How many servers I need? I know it's difficult to say something like
that, but right now, I have 2 servers, 4 core and 7.5 go RAM each one.
It's really slow.

Any idea?

Thanks

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

Hi,

Yes that's why I open a new topic here : Redirecting to Google Groups
There is much more informations.

I succeeded to have queries about 2-3 seconds, nothing better :frowning: My index of 15 millions docs must have a problem somewhere I think.

I use facets but with search_type 'count' and only facet_filter, not query…. Still slow…

Any idea to optimise ?

Thanks,
Regards,

Fabien

On Sep 18, 2013, at 11:47 PM, Alexander Reelsen alr@spinscale.de wrote:

Hey,

you havent said anything about your configuration or you definition of slow and even your expectations. Maybe you can elaborate about all this bit more...

Generally the best thing is to start with one server and see when it degrades in performance (this might be indxing speed, query latency or whatever metric you have defined for yourself) and then start slowly with scaling out.

--Alex

On Wed, Sep 18, 2013 at 12:09 PM, Fabien Guiraud fabien1909@gmail.com wrote:
In fact, for users who knows AWS, I'm on EC2, with m1.large instances

On Wednesday, September 18, 2013 12:05:18 PM UTC+2, Fabien Guiraud wrote:
Hi all,

I need to know how scale ES.
I have 15millions docs for one index, 5millions for another one, and I do a lot of big faceting (terms stats, etc...).

How many servers I need? I know it's difficult to say something like that, but right now, I have 2 servers, 4 core and 7.5 go RAM each one.
It's really slow.

Any idea?

Thanks

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

--
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/_mlplReadmw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to elasticsearch+unsubscribe@googlegroups.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.
For more options, visit https://groups.google.com/groups/opt_out.