Thanks Ed!

I have a question about load balancing that I can't find in the documentation, and I need for a project.
We use ICN to view data stored in CMOD, and use the round-robin approach to our servers for load balancing.

(We store the data currently in DB2 databases.)

My question is - Does ICN need persistence load balancing?   Or is the round-robin method for logging into ICN handling those potential needs.

Any assistance would be appreciated.

For us (and our high retentions) we would do #3 or even by year, but that's a matter of opinion.
I'd love to see the code you use, as I am currently setting up a similar process and you could save me some time.

thanks.  We don't get that, or any error, it just won't allow me to map more AGs to the folder.
But, to give you some "just curious" info...
Most of our folders are much smaller than this, but on the "high" side, we have a handful of folders that have around 15,000 apps in them, but they aren't "viewed".  Those folders are there just to see if something loaded (looking at results), when there is an issue; they contain data that is sent elsewhere and is not viewable in CMOD.
We have a handful that are in the 10,000 applications range, but this is based on the needs of the customer, and they prefer the possible hit in performance over having to look in more than one folder.

Thanks.  Yes, we know about the potential performance issues, but our environment and customer base require the setup we have chosen.  We have applied other processes to reduce the hit, such as a default timeframe in the selection criteria, which our customers know that increasing the date range can decrease the response time.

We are on 9.5, but are in the process of upgrading to version 10.  We hope to have the upgrade completed by the end of the month.

I will create a PMR for this, so we can determine if it's a bug in the system or our install. 

Thank you, again for your quick response. 
Janine

Is there a maximum number of application groups you can map within a folder?  I am asking because we have a folder that has a large number of application groups in it, most of which only select specific applications within the application group, and I keep getting an error when I try to map additional AGs to the folder.  Although I am selecting the correct field and AG from the list, when I click on ADD, I get an error saying that it has already been mapped, but it shows me a different AG then what I selected.
I've searched the manuals and this site for an answer, and can't find anything.
I can, of course, split the folder out, but that would require additional information and training updates for all of my customers.

EXAMPLE:
Am trying to map fields for AG CORP_CST_I00_R3654_001.
Select database name and for that field and AG from the Application Group box (Note:  none of the fields for that AG have been mapped)
Receive error:  A folder fields with name xxx has already been mapped to the selected field in application group DMG_DMM_I00_R7400_002
Yes, it has already been mapped to THAT AG, but that's not the AG I'm trying to map it to.

Any help would be greatly appreciated.
Thanks,
Janie

Ok, I found out what causes that.  We couldn't use the DSID because the way it was originally set up in the application was not getting the entire number.  We have changed that.  We still have a few that start 1 character off in the JES data, but will probably use a floating trigger and/or index to use that - or - we will just expand the data size and use the option to remove spaces at the beginning and possibly the period that shows up in the end.
Thank you all for your input!

When we load our JES output to CMOD, using ARSYSPIN, and then view the output (in the Web viewer, the windows client OR the CICS viewer), it does not list the output in the same (step) order as the job was run OR how it is listed in the JES output queue.
Since many steps run within the same second, and the job's steps are not in an alpha-numeric order, we can't do a sort on the fields in the folder either.

Can you tell me how to get the output to display in the viewers in the same order that it is created and listed in the JES output queue?

Thanks
J9

Thanks Ed.  THis is good info to know.

Thanks for the quick reply Ed. 

We are using ARSRPSUB, but we can't have the jobname removed, allowing ARSODF to run them, because the high priority ones we need have to have the jobname for the Print room to recognize and process them. 
Am I able to select which ones use the ARSODF Proc, and which don't?  If so, I could try changing the ones that are NOT high priority to the ARSODF Proc, because we don't need the jobname on them, that just makes our lives easier.  And if I do that, will it then process all the Low Priority first, since ARSODF Proc is processing them, causing additional delays?

We have a LOT of data going through ODF.  There are specific types (desitinations, recipients, etc) that we need to have processed at a higher priority than others.
Does anybody know of a way to set a priority on ODF processing?

Scenario:  5 high-priority sets are sent to ODF, then 5000 low-priority sets are sent, then another 5 high-priority sets are sent.  We can't wait for the 5000 in the ODF queue to process before it gets to the next set of high priority. 
We can't use any of the "time of day", etc, options to change the order of the sets that are sent to ODF but need to have the high-priority sets in the queue processed first.
Other than starting another ARSODF task to handle them, is there anything we can set in the system or in the distributions to indicate the priority for ODF to process them?

Thanks,
Janine

Thanks scottnys.
Your shop sounds a lot like ours.
Yes, we have customized exits for a all the reasons you mentioned, plus a few more.  From what I read in the manuals, etc, I figured it was an issue with the exit not passing the information to the manifest exit, but was trying to avoid having to change our exits.  We have 4 different exits to handle the scenarios we need, and changes to our Cobol programs take several weeks to get implemented in Production. 
I have put this issue on our back burner, to re-visit when we are more stable, and am using the work around in the meantime.  All the details you supplied will be helpful in my documentation - thanks for taking the time to respond!

May not be the best solution, overall, but I found a work-around for us.
(In short) since the output I was working with was always the same size, I set up our email system to accept a specific number of lines for them, which forces a new page before the manifest.
Thanks for your suggestion, though!

Yes, we have that PTF applied.  I guess it doesn't fix the problem for us.

Thanks Greg, I have my Production Control resource checking now.  I'm pretty sure we have the same "cause" as you; I'll let you know.