Messages

1741

Report Indexing / Re: Intermittent Index and Load Failures arsadmin.c

« on: February 08, 2012, 09:07:06 AM »

Looks like a transient database issue.  Check db2diag.log (generally /home/archive/sqllib/db2dump/db2diag.log) for hints as to what the problem might be.

-JD.

1742

Other / Re: ARSADMIN vs ARSMAINT

« on: February 07, 2012, 08:34:27 AM »

It might make more sense to duplicate the AG/App/Folder and just delete the old one.  If you need to keep some data from that group, use 'arsdoc get' to fetch the data you want, and load it into the newly defined AG. 

Once the old group is deleted, you can rename the new one to match the name of the old one, and everything goes back to normal. 

Let us know how it goes!

-JD.

1743

MP Server / Re: Problem running arssockd after db2 restore.

« on: February 07, 2012, 08:26:38 AM »

What commands did you use to restore the database?  Which user ID/ instance did you restore it into?

I'm guessing that you restored the QA database into the same instance as your dev instance, which isn't likely to work.

If you want to change the schema name, you're going to need to export all the tables in IXF format and import them into the new instance and database.

-JD.

P.S.  Alessandro is right -- you're WAY too far out of support.  Get upgraded as soon as you can!

1744

Other / Re: Application groups and report distribution

« on: January 31, 2012, 06:38:25 AM »

Fewer Application Groups with more Applications are better.  Each Application Group you 'bind together' at the Folder level is a separate query -- why do 25 queries when you can do just one?.  If it supports your needs, one AG with 200 Apps is the way to go.  You can use the AppID field to filter/restrict your results at query time.

If your volume is high, increase the max rows per table to equal your approximate monthly volume, plus a multiple of your expected annual growth.  (So, 55 million documents a month with 10% annual growth = 60 to 70 million rows per database table.)  This will help keep performance linear as your system grows.

-JD.

1745

Other / Re: Application group with no storage set

« on: January 27, 2012, 06:32:53 AM »

My personal 'Best Practices' when it comes to storage are this:

1)  Every Application Group gets it's own Storage Node (preferably by the same name), and is configured to 'migrate later' (advanced tab).
2)  Every Storage Set is configured for TSM -- even if you have no intention of using TSM -- because "the only thing that stays the same is change".
3)  Use TSM.  It's the only way to have a truly optimized system.  Always move unused data to tape ASAP, to make more room for the data people ARE accessing.

1746

Report Indexing / Can you recommend a tool for creating AFP Resources, like FormDefs?

« on: January 26, 2012, 07:35:26 AM »

Hi everyone.

I have a customer who is looking to create AFP data streams from text output from an AS/400.  Creating an overlay (ie, the pre-printed form) from a TIFF is straightforward -- but creating the other components like FormDefs and PageDefs seems to be a little more complicated.

What utilities do you use to create these resources?

-JD.

1747

MP Server / Re: ARSLOAD question

« on: January 24, 2012, 06:06:22 AM »

Hi.

Don't rename the .out file -- leave it where it is, just issue the 'touch' command with the same base filename and the .ARD extension.

So, in your case, TestApp.0.TestAppGroup.TestApp.out becomes TestApp.0.TestAppGroup.TestApp.ARD.

If you're not familiar with the UNIX 'touch' command, it creates a zero-byte file if it doesn't exist, or updates the access timestamp if the file DOES exist.

If you need to do a LOT of these, you can automate it easily with a few lines of ksh/bash:

Code: [Select]

for i in *.out
do
  touch ${i%.out}.ARD
done


1748

Documentation / Re: How to upload a document to test from OD client

« on: January 19, 2012, 08:49:19 AM »

Hi Hrushi.

Loads are performed at the command line.  You'll need to open a command window and run the 'arsload' command.  You really need to read the manuals to get a good idea of how this works.

-JD.

1749

Announcements & News / Re: Should we create a LinkedIn Group?

« on: January 17, 2012, 03:17:42 PM »

The ODUG Group on LinkedIn has been made available to the public -- however, this will remain the official ODUG Forum.

The LinkedIn Group can be found here:

http://www.linkedin.com/groups?gid=4148818

-JD.

1750

z/OS Server / Re: how to correct for ARS0131E, no logical place to retrieve object ?

« on: December 28, 2011, 03:41:33 PM »

Hi Larry.

On the UNIX platforms, this would mean that the ars.cache configuration file had no cache filesystems listed, and/or the TSM configuration files (dsm.opt/dsm.sys) were missing or incorrect.

Not sure how this is configured on z/OS, but I hope it's a direction to go in, at the very least. 

1751

Other / Re: Tablespace Creation exit Routines

« on: December 28, 2011, 03:37:06 PM »

Hi Gopi.

Maybe the code you pasted in was just an example -- but why are you trying to re-write the table deletion code?  The sample you provided doesn't appear to do anything other than delete the same table over and over again (regardless of which table was requested to be deleted).

CMOD does an excellent job of managing tables all by itself -- I'm having a hard time trying to imagine a situation where you'd want to modify the table deletion step.

Can you fill us in on what goal you're working towards?

-JD.

1752

MP Server / Re: Is it necessary to have the EXIT programs in C?

« on: December 21, 2011, 05:07:20 AM »

It might be possible, but I haven't seen any examples, and it would likely be extremely slow compared to C.  I'm not very familiar with Java, but wouldn't you need to load an instance of the JVM each time?  That would slow down your retrievals dramatically.  If there were a large number of accesses in a short period of time, you would not only slow down the server while it loads up multiple JVMs, I can imagine that it would be easy to use up all the RAM in the server very quickly.

-JD.

1753

OD/WEK & JAVA API / Re: Modify output using ODWEK

« on: December 21, 2011, 04:52:21 AM »

Hi Pankaj.

You should be able to accomplish what you're asking for with an Output exit.  And as far as I know, it must be written in C.  C may be complicated, but it results in an extremely fast finished product.

-JD.

1754

MP Server / Re: CMOD Warning after Successful Loads

« on: December 20, 2011, 08:58:44 AM »

It sounds like you don't have permissions (or the filesystem is full) in a temporary directory like /tmp or /arstmp or wherever you've configured your CMOD temporary directory.

-JD.

1755

iSeries / Re: Retrieve a CMOD document and email found document

« on: December 16, 2011, 06:53:29 AM »

I know that you're using iSeries, but hopefully the process for sending an eMail from the server is similar to UNIX/Linux.

First, once you retrieve the file, it needs to be in a format that the customer can view it in.  For most companies, this is PDF.  You could also convert to GIF, JPEG, or BMP if those options are desirable.  Once in the PDF format, you'll need to convert it so that it can be sent in an eMail - normally using MIME base64 ( http://en.wikipedia.org/wiki/Base64 ).  After that, format your eMail ("Dear customer, here's your invoice.") and append the converted file.

It will take some trial and error, but it's entirely possible.

The only other thing to be aware of is that your mail server needs to be properly configured.  If I had a dime for every time an automated eMail is rejected from my server (because the remote server is misconfigured, and looks like a spammer) I'd be eating a steak dinner every night!

Good luck!

-JD.