| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| WMB V8 runtime Uncompiled Code |
« View previous topic :: View next topic » |
| Author |
Message
|
| lancelotlinc |
 Posted: Tue Sep 18, 2012 10:13 am Post subject: WMB V8 runtime Uncompiled Code |
 |
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
Now that an official way to deploy assets to WMB runtime without compiling them first has been released (ie. mqsipackagebar), what is the behavior of the runtime between compiled and uncompiled code?
Does uncompiled code get compiled at deploy time by the mqsideploy command? Or, is the code run at runtime like would be from an interpreter?
For message flows with many ESQL lines, would there be latency differences between deployed resources that were built by mqsicreatebar or by mqsipackagebar ?
BTW this is not a critique, I like the idea of packaging resources. I'm interested to know what the expected performance differences are.
Another angle: mqsicreatebar resources would throw errors at deploy time if variables used are not in scope. Is the same syntax checking done at deploy time for packaged resources?
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
 |
| NealM |
| Posted: Wed Sep 19, 2012 10:09 am Post subject: |
|
|
Master
Joined: 22 Feb 2011
Posts: 230
Location: NC or Utah (depends)
|
Not an answer but,
Yeah, I'd like to see the Hursley WMB development team put out some comparison numbers in their performance reports when there is more than one way to skin a rabbit. Stuff like compiled vs uncompiled, DFDL vs MessageSet in a mapping node, Broker-wide HTTP listener vs EG listener, etc. I've asked for a few, no response. Why don't you email Tim, maybe if he/they hear it from more than just me.... (Their reports do say they want feedback and give email addresses) |
|
| Back to top |
|
|
| kimbert |
| Posted: Thu Sep 20, 2012 1:05 am Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5542
Location: Southampton
|
OK, I'll bite on this one...
| Quote: |
| DFDL vs MessageSet in a mapping node |
I assume you mean 'DFDL vs MRM'. Don't confuse 'message set' with 'MRM' because a message set can and should be used in just about any v6/v7 message flow. For most types of data format DFDL outperforms MRM quite comfortably.
The mapping node has no effect whatsoever on parsing performance. Nor does ESQL. Nor Java/PHP/.Net. All of those transformation languages operate on the message tree, but none of them are involved in creating the message tree. I'm sure you know that, but other readers might not. |
|
| Back to top |
|
|
| NealM |
| Posted: Thu Sep 20, 2012 8:49 am Post subject: |
|
|
Master
Joined: 22 Feb 2011
Posts: 230
Location: NC or Utah (depends)
|
Yes kimbert, that is what I meant. And I should mention that in response to an email, Hursley did respond. In part (related to Mapping Nodes and DFDL):
| Quote: |
| I cannot say whether there is a good return on investment for <my company> to redevelop existing message flows - that depends on multiple factors like cost of development/testing. complexity of the flows etc. Maybe they could pick the high volume ones? |
It would just be nice to have some published performance deltas to flash before the pursestring holders. But the suggestion in the quote is the approach we will be using. |
|
| Back to top |
|
|
|
|
|
| |
|
Page 1 of 1 |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
