You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@ctakes.apache.org by "Baas,Leah" <Le...@SanfordHealth.org> on 2019/01/31 16:43:51 UTC
cTAKES heuristic for memory/CPU allocation
Hi everyone,
I’m setting up a new VM to process ~13,500 files using the cTAKES default clinical pipeline. Looking for a good heuristic for RAM/CPU allocation. Can cTAKES take advantage of multiple CPUs? Or is it designed only to utilize a single thread?
Thanks for the help.
Leah
-----------------------------------------------------------------------
Confidentiality Notice: This e-mail message, including any attachments,
is for the sole use of the intended recipient(s) and may contain
privileged and confidential information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply e-mail and destroy
all copies of the original message.
Re: [EXTERNAL] Re: cTAKES heuristic for memory/CPU allocation
Posted by "Baas,Leah" <Le...@SanfordHealth.org>.
Thanks a lot, Mike!
Leah
From: Michael Trepanier <mi...@metistream.com>
Reply-To: "user@ctakes.apache.org" <us...@ctakes.apache.org>
Date: Thursday, January 31, 2019 at 11:44 AM
To: "user@ctakes.apache.org" <us...@ctakes.apache.org>
Subject: [EXTERNAL] Re: cTAKES heuristic for memory/CPU allocation
The cTAKES Default Processing Pipeline requires about a minimum of 3G of RAM due to the size of the embedded HSQLDBs (that is the default). However, providing a fair bit of overhead is generally a good idea.
As for multi-threading, I have been using the ThreadSafeLvg class. Per the component-use guide:
"cTAKES was not originally designed to be thread safe.
If you would like to experiment with making it thread safe, see class ThreadSafeLvg in ctakes-lvg."
https://cwiki.apache.org/confluence/display/CTAKES/cTAKES+4.0+Component+Use+Guide
Mike
On Thu, Jan 31, 2019 at 8:44 AM Baas,Leah <Le...@sanfordhealth.org>> wrote:
Hi everyone,
I’m setting up a new VM to process ~13,500 files using the cTAKES default clinical pipeline. Looking for a good heuristic for RAM/CPU allocation. Can cTAKES take advantage of multiple CPUs? Or is it designed only to utilize a single thread?
Thanks for the help.
Leah
-----------------------------------------------------------------------
Confidentiality Notice: This e-mail message, including any attachments,
is for the sole use of the intended recipient(s) and may contain
privileged and confidential information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply e-mail and destroy
all copies of the original message.
--
[etiStream Logo - 500]
Mike Trepanier| Senior Big Data Engineer | MetiStream, Inc. | mike@metistream.com<ma...@metistream.com> | 845 - 270 - 3129 (m) | www.metistream.com<http://www.metistream.com/>
Re: cTAKES heuristic for memory/CPU allocation
Posted by Michael Trepanier <mi...@metistream.com>.
The cTAKES Default Processing Pipeline requires about a minimum of 3G of
RAM due to the size of the embedded HSQLDBs (that is the default). However,
providing a fair bit of overhead is generally a good idea.
As for multi-threading, I have been using the ThreadSafeLvg class. Per the
component-use guide:
"cTAKES was not originally designed to be thread safe.
If you would like to experiment with making it thread safe, see class
ThreadSafeLvg in ctakes-lvg."
https://cwiki.apache.org/confluence/display/CTAKES/cTAKES+4.0+Component+Use+Guide
Mike
On Thu, Jan 31, 2019 at 8:44 AM Baas,Leah <Le...@sanfordhealth.org>
wrote:
> Hi everyone,
>
>
>
> I’m setting up a new VM to process ~13,500 files using the cTAKES default
> clinical pipeline. Looking for a good heuristic for RAM/CPU allocation. Can
> cTAKES take advantage of multiple CPUs? Or is it designed only to utilize a
> single thread?
>
>
>
> Thanks for the help.
>
>
>
> Leah
>
> -----------------------------------------------------------------------
> Confidentiality Notice: This e-mail message, including any attachments,
> is for the sole use of the intended recipient(s) and may contain
> privileged and confidential information. Any unauthorized review, use,
> disclosure or distribution is prohibited. If you are not the intended
> recipient, please contact the sender by reply e-mail and destroy
> all copies of the original message.
>
--
[image: MetiStream Logo - 500]
Mike Trepanier| Senior Big Data Engineer | MetiStream, Inc. |
mike@metistream.com | 845 - 270 - 3129 (m) | www.metistream.com