LAE

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: xathras

Hi All,

Hoping there is a document somewhere explaining all the different variables settings in the prop files in the conf directory. e.g. ls_brain.prop

The reason I ask, is we recently found that our 4.5.4 Installation was running nodes in sequential order (e.g. one node at a time), rather than running statements in parallel.
As an example, we have a composite node that runs one node to get the maximum number from the database (lets say subscription number), then a filter node has four outputs, splitting a min number and maximum out (e.g. max number is 1000. Output 1 range 0-249, Output 2 range 250-499, Output 3 500-749, Output 4 750-1000) each output is then pinned to a JDBC SQL Node querying data been those ranges. Since each node is not clocked and independent they should run concurrently.

We originally thought maybe the thread pooling we implemented was causing this so we removed the config, restarted and nothing changed. We then noticed the controller was running in default so changed that to streaming, restarted and that didn't help, we checked nothing else was running on the server causing this and then we then restored a config file from 4.1.8 install (that the Lavastorm Team did for us) and found the following attribute:

ls.brain.controller.farm=( "localhost:${ls.brain.server.port}:16" ) - in the default 4.5.4 install this was 1. By changing it to 16, we noticed things running in parallel.

So what does that setting mean along with the others?

Wayne

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: Tim Meagher

Hi,

Are you running this as a BRX?
If so, when running as a BRX, the format for the ls.brain.controller.farm property is:
(<host>:<port>:<numberOfThreads>)
If you have more than one server to use, then you can comma delimit the servers within the brackets.

Some further discussion of this topic can be found here:
http://community.lavastorm.com/showt...nController-py

So, without specifying this, the number of threads I believe is 1, which means there will be no parallelism.
If you are running through BRE, rather than as a BRX using the command line controller, then you can specify the number of threads in your server farm configuration under Tools->Preferences->Server Farms.
For running as a BRX you need to either modify this property in ls_brain.prop or pass this as an argument to the controller when you run it.

Unfortunately, there isn't a document which describes the function of all of the various options in ls_brain.prop at the moment.
This has been raised as issue # 4184 in our bug tracking system in order to provide the necessary documentation.
Until this issue is resolved, if you have similar questions relating to the operation and format of such properties, someone within the forums should be able to answer relatively quickly.

Regards,
Tim.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: xathras

Thanks for the update Tim. I will check it out.

I can confirm this is when we're running it as a BRX

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: anjuls

Hi Guys,
I am new to Lavastorm software and I am constantly facing a problem with brain temp directory taking large amount of space. Is there any way to control it? Right now, I am deleteing files older than one day. I am more interested in exploring an option of putting a parameter in ls brain config file to control the max temp size or automatic purge.
Please advise.

Thanks in advance,
Anjul

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: xathras

Best way is to have a script setup to look for files older than x number of days old in the temp directory and purge and then have it configured in crontab. Let me see if I can dig out ours and post.

Wayne

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: spalluru

Hi Tim,
Just wanted to know is this document now available , if so can you please share the same ?
-Sreekanth

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: Tim Meagher

Hi,

There is still no document available describing these parameters and the enhancement request is still open.

When it becomes available it will no doubt be either installed with the software or uploaded to the documentation site linked from the software, or both and will be called out in the release notes.

I'll try to remember to update this post when such a document becomes available.

Regards,
Tim.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: anjuls

Hi Tim,

Is this document now available? We keep on struggling to understand how those properties can be used or modified to tune lavastorm in our environment. Hoping to get this soon. It would be very helpful.

thanks
Anjul