-Mm Startup Parameter...That old chestnut!

Hi everyone, after suffering some performance issues the other day I thought it was worthwhile revisiting the Performance Tuning guide. I noticed that the example pf file in that document has a -Mm value of 10240, All of our template .pf files in our config directory have a value of 1024, which appears to be the default when I look at the Progress website.

The document does not mention the -mM value at all or give an explanation, although it does for other parameters.

I am hoping someone can clarify.

9.05.701 SQL 64 Non-Unicode. (Yes - Non Unicode!)

Hi Hally,

Here is what you are looking for:

Best Regards,
Carlos Q.

1 Like

Thanks Carlos, I’d read that earlier. What Specifically I was wanting to know what why does the Epicor document show an example of -Mm as 10240 but the defaults in the template files are 1024…

The comment around “If your network works more efficiently with smaller messages then reduce -Mm” How do you know, has anyone done any tests around this they would like to share.



I think the best is that you have a look at the OpenEdge documentation mentioned on the article “Startup Command and Parameter Reference Guide Documentation”.

In fact, just by looking at the extract from that document they include there, it makes sense to increase it because we all know Epicor db records are large.

Here some examples:
-Mm 4096 -mmax 65534 -Bt 20480 <-- one I found on epicor answerbooks about perf.
-Mm 1024 -mmax 65534 -Bt 16384 <-- the default
-Mm 10240 -mmax 65534 -Bt 40960 <-- this is the one you mention.

All are valid values since:
The -Mm sets the maximum message size in bytes.
The multi-user default value is 1024.
min 350 and max 32600 bytes.

Best practices are an evolving thing, and in the pre-E10 days we cut physical media for the base install so we couldn’t easily change things like base templates after it was in the wild even if we discovered a new best practice because they were delivered with the base install. In E10.1+ that is no longer a problem–as we discover new settings that are within our ability to adjust and it wouldn’t potentially affect other products we can do so at the release or patch levels.

Not sure if that was your actual question, but, that is the existential reason why that document exists at all and why some settings in that document are different than the defaults.

Thanks Nathan, I understand totally. Next question which is right? If I can hear it from the horses mouth, then I be a happy person.

Everything else being equal, most customers would see the most performance benefit by using the settings in the tuning guide (10240).


Thanks for the clarification Nathan.