TM1 10.2.2 FP6

User avatar
Steve Rowe
Site Admin
Posts: 2410
Joined: Wed May 14, 2008 4:25 pm
OLAP Product: TM1
Version: TM1 v6,v7,v8,v9,v10,v11+PAW
Excel Version: Nearly all of them

Re: TM1 10.2.2 FP6

Post by Steve Rowe »

Hi, As mentioned above IFs are only available via an explicit support request / PMR to IBM, you won't see them on the general download site
Technical Director
www.infocat.co.uk
Jefflinde
Posts: 12
Joined: Wed May 25, 2016 2:28 pm
OLAP Product: TM1
Version: 10.2.2
Excel Version: 2013

Re: TM1 10.2.2 FP6

Post by Jefflinde »

Sorry, thank you for the clarification.
TomaszB
Posts: 19
Joined: Thu Mar 21, 2013 9:00 am
OLAP Product: TM1
Version: 9.5.2 10.1.1 10.2.2
Excel Version: 2003 2010

Re: TM1 10.2.2 FP6

Post by TomaszB »

Hi all,

I received last the information that after installing 10.2.2 FP6 (even IF2) persistent feeders can delete cube data.

http://www-01.ibm.com/support/docview.w ... wg1PI70620

After asking IBM when it can really happen, I got the answer that the data that could get lost are those that fulfill three conditions :
1. datapoints not actually rules-derived
2. but holding stored values
3. and targeted by an actually triggered FEEDER (aka actually get fed).
(Note it doesn't matter whether the FEEDERS are persistent or actual FEEDER statements are executed (refer PI37250 ).)

...as best practices are to have not at all FEEDERS targeting on non-Rules-derived datapoints it could be that some 'small' projects are not affected, but I would never say that there are not overfeeds or they will not be build in in the future.

Have someone experienced problem with the feeders so far?

Thanks
Tomasz
Last edited by TomaszB on Fri Oct 21, 2016 9:05 am, edited 1 time in total.
User avatar
Steve Rowe
Site Admin
Posts: 2410
Joined: Wed May 14, 2008 4:25 pm
OLAP Product: TM1
Version: TM1 v6,v7,v8,v9,v10,v11+PAW
Excel Version: Nearly all of them

Re: TM1 10.2.2 FP6

Post by Steve Rowe »

Thanks for alerting everyone Tomasz, that's a pretty bad bug, though feeders that are a different shape to the rules are a little unusual, though not rare.

Maybe they'll rush a fix out including working SSL certs in the core build.
Technical Director
www.infocat.co.uk
lotsaram
MVP
Posts: 3651
Joined: Fri Mar 13, 2009 11:14 am
OLAP Product: TableManager1
Version: PA 2.0.x
Excel Version: Office 365
Location: Switzerland

Re: TM1 10.2.2 FP6

Post by lotsaram »

TomaszB wrote:Hi all,

I received last the information that after installing 10.2.2 FP6 (even IF2) persistent feeders can delete cube data.

http://www-01.ibm.com/support/docview.w ... wg1PI70620

After asking IBM when it can really happen, I got the answer that the data that could get lost are those that fulfill three conditions :
1. datapoints not actually rules-derived
2. but holding stored values
3. and targeted by an actually triggered FEEDER (aka actually get fed).
(Note it doesn't matter whether the FEEDERS are persistent or actual FEEDER statements are executed (refer PI37250 ).)

...as best practices are to have not at all FEEDERS targeting on non-Rules-derived datapoints it could be that some 'small' projects are not affected, but I would never say that there are not overfeeds or they will not be build in in the future.

Have someone experienced problem with the feeders so far?

Thanks
Tomasz
This is indeed a bad bug. It isn't an FP6 issue though. We have it reported on 10.2.2 FP5 under PMR 07689999616
Attached is the test case submitted to IBM support to reproduce on sData.
Attachments
Missing TM1 Data.docx
(281.89 KiB) Downloaded 350 times
Please place all requests for help in a public thread. I will not answer PMs requesting assistance.
User avatar
Steve Rowe
Site Admin
Posts: 2410
Joined: Wed May 14, 2008 4:25 pm
OLAP Product: TM1
Version: TM1 v6,v7,v8,v9,v10,v11+PAW
Excel Version: Nearly all of them

Re: TM1 10.2.2 FP6

Post by Steve Rowe »

Quick post, Friday afternoon and all that. I've reproduced this bug in FP6 vanilla but have not been able to do so in IF1. I've not been able to track down definitive confirmation from IBM anywhere that the APAR was fixed in IF1.

Many thanks for posting the use case LotsaRam, much appreciated.

Cheers and bonne weekend!
Technical Director
www.infocat.co.uk
TomaszB
Posts: 19
Joined: Thu Mar 21, 2013 9:00 am
OLAP Product: TM1
Version: 9.5.2 10.1.1 10.2.2
Excel Version: 2003 2010

Re: TM1 10.2.2 FP6

Post by TomaszB »

Hi All,

Many thanks for posting the use case LotsaRam and for the testing from your site Steve.
I was able to reproduce it using the FP6 but I do not have the IF1.

It would be good it the IF would fix it.
I will request it from IBM and post my results later.

Cheers!
Tomasz
TomaszB
Posts: 19
Joined: Thu Mar 21, 2013 9:00 am
OLAP Product: TM1
Version: 9.5.2 10.1.1 10.2.2
Excel Version: 2003 2010

Re: TM1 10.2.2 FP6

Post by TomaszB »

TomaszB wrote:Hi All,

Many thanks for posting the use case LotsaRam and for the testing from your site Steve.
I was able to reproduce it using the FP6 but I do not have the IF1.

It would be good it the IF would fix it.
I will request it from IBM and post my results later.

Cheers!
Tomasz

Hi all,

IBM just released the IF4 (10.2.5260.1014) for 10.2.2 FP6 which includes the fix of the PI70620 Feeders problem.
I tested it and the problem cannot be reproduced anymore!

Cheers
Tomasz
MSidat
Community Contributor
Posts: 110
Joined: Thu Aug 26, 2010 7:41 am
OLAP Product: TM1, PA
Version: PAL 2.0.8
Excel Version: 2016
Location: North West England

Re: TM1 10.2.2 FP6 Now up to IF6

Post by MSidat »

FP6 is now up to IF6 if anyone requires any further fixes, which would include fixes for the following, just got given this to fix the potential loss of date due to feeders issue:

PI68108 COMMAND CENTER REST API CONNECTION CRASHES TM1 SERVER
PI70584 MACRO IN XLSM FILE IS LOST AFTER UPLOAD TO TM1WEB AND SLICED (EXPORTED) TO *.XLSX FORMAT
PI70956 PERFORMANCE ISSUES RUNNING PARALLEL TI PROCESSES WITH TM1RUNTIAFTER UPDATE FOR FIXPACK 6
PI71503 TM1 server crash
Always Open to Opportunities
lotsaram
MVP
Posts: 3651
Joined: Fri Mar 13, 2009 11:14 am
OLAP Product: TableManager1
Version: PA 2.0.x
Excel Version: Office 365
Location: Switzerland

Re: TM1 10.2.2 FP6

Post by lotsaram »

Do you have the fix ID? The search functionality on Fix Central is so woeful that's usually the only way to find something.
Please place all requests for help in a public thread. I will not answer PMs requesting assistance.
MSidat
Community Contributor
Posts: 110
Joined: Thu Aug 26, 2010 7:41 am
OLAP Product: TM1, PA
Version: PAL 2.0.8
Excel Version: 2016
Location: North West England

Re: TM1 10.2.2 FP6

Post by MSidat »

Is this what you are after?

10.2.2.6-BA-CTM1-IF1016

http://www.ibm.com/support/fixcentral/s ... login=true

The link itself does not work for me as it comes up with a "Restricted Release" type of message. I have downloaded the fix, but it came via our IBM Partner. You may need to specifically request it via IBM.
Always Open to Opportunities
AmbPin
Regular Participant
Posts: 173
Joined: Sat Mar 20, 2010 3:03 pm
OLAP Product: TM1
Version: 9.5.2
Excel Version: 2007-10

Re: TM1 10.2.2 FP6

Post by AmbPin »

qml wrote:considering that it is possible to check in the Prolog if your source view will return at least one value
Please could you tell me how you can check?
AmbPin
Regular Participant
Posts: 173
Joined: Sat Mar 20, 2010 3:03 pm
OLAP Product: TM1
Version: 9.5.2
Excel Version: 2007-10

Re: TM1 10.2.2 FP6

Post by AmbPin »

qml wrote:considering that it is possible to check in the Prolog if your source view will return at least one value
Please could you explain how you might do this?
I have seen a technique used with a generic process that sets the data view and initializes a counter in the prolog, in the data tab the variables are not referenced but the counter incremented. If the counter is incremented to one, jump to the epilog and the view returns at least one row.
Is that the sort of thing you mean?
User avatar
Alan Kirk
Site Admin
Posts: 6606
Joined: Sun May 11, 2008 2:30 am
OLAP Product: TM1
Version: PA2.0.9.18 Classic NO PAW!
Excel Version: 2013 and Office 365
Location: Sydney, Australia
Contact:

Re: TM1 10.2.2 FP6

Post by Alan Kirk »

AmbPin wrote:
qml wrote:considering that it is possible to check in the Prolog if your source view will return at least one value
Please could you explain how you might do this?
I have seen a technique used with a generic process that sets the data view and initializes a counter in the prolog, in the data tab the variables are not referenced but the counter incremented. If the counter is incremented to one, jump to the epilog and the view returns at least one row.
Is that the sort of thing you mean?
It's the only way you could do it (the return being done by global variables). The whole of the Prolog executes before even one row of data is read. It has to because the data source may be set in the Prolog tab so the Prolog may not even know what the data source is until its very last line.

However, that comes with one big-backsided old caveat.

If the data source returns no records, the process will work fine. I know, I've done it. But if the data source returns nothing because it cannot be opened... the chore / process chain will abort because the TI language, having had 3/10ths of sod-all of gorramn improvements done on it in the last, what, 18, 19 years, still has no adequate error handling capabilities. {Slaps hand to head} What am I saying? It has no freaking error handling capabilities at all. 18 sodding years, no On Error Goto, and fuhgeddabout Try / Catch, that's out there in fantasy land.

For a text file it's easy to work around this and prevent your chore from crashing and burning by testing for the existence for the file before you ever try to access it. For something like an ODBC connection this is more of a problem. Same with a cube view which may be corrupt from, say, an empty subset.

So while the "call another process to check the presence of records" approach is a viable option... it's one that only works if the data source is valid.

Edit: I should mention though; in 10.2 upwards there is also the option of, instead of calling another TI process, calling some Java code to check the data source and whether it has rows. Java does of course actually have error handling, so it offers a potential way around that problem. It ain't pretty and it stinks up the documentation and maintenance of your code, but it'll work.
"To them, equipment failure is terrifying. To me, it’s 'Tuesday.' "
-----------
Before posting, please check the documentation, the FAQ, the Search function and FOR THE LOVE OF GLUB the Request Guidelines.
AmbPin
Regular Participant
Posts: 173
Joined: Sat Mar 20, 2010 3:03 pm
OLAP Product: TM1
Version: 9.5.2
Excel Version: 2007-10

Re: TM1 10.2.2 FP6

Post by AmbPin »

Thanks Alan,

I just built a test to try the call to another process. I started down this route because of the new FP6 bug, mentioned in this thread, that crashes the service if you dynamically set an empty source view.

I use a generic TI to create my views, so I have amended that to:-
* Set the view created in the prolog as a data source and initialize a counter;
* In the data tab, if we encounter a value then increment the counter and call ProcessBreak;
* In the Epilog if the counter is One then there are some values in the view;

I tend to use a user/process/... cube to pass back values instead of global variables but I think the principle is the same.

What I have found though is that although the generic process above does ignore all of the variables and has no reference to any of the variables in the data tab, I still have to include DisableMTQViewConstruct() in the prolog to stop the service falling over.

Since in my case, this is a work-around for an FP6 bug I am not sure it is any better than putting DisableMTQViewConstruct() in the calling process, the one that calls the generic view construct and that actually loops through the view data.

Hope that makes sense!
lotsaram
MVP
Posts: 3651
Joined: Fri Mar 13, 2009 11:14 am
OLAP Product: TableManager1
Version: PA 2.0.x
Excel Version: Office 365
Location: Switzerland

Re: TM1 10.2.2 FP6

Post by lotsaram »

As I recall this bug had something to do with the introduction of temp objects and there is a TI function which you can declare in the prolog before the view creation which avoids the crash. Sorry can't remember what it was off the top of my head.
Please place all requests for help in a public thread. I will not answer PMs requesting assistance.
AmbPin
Regular Participant
Posts: 173
Joined: Sat Mar 20, 2010 3:03 pm
OLAP Product: TM1
Version: 9.5.2
Excel Version: 2007-10

Re: TM1 10.2.2 FP6

Post by AmbPin »

Was the function you are thinking of "DisableMTQViewConstruct"
lotsaram
MVP
Posts: 3651
Joined: Fri Mar 13, 2009 11:14 am
OLAP Product: TableManager1
Version: PA 2.0.x
Excel Version: Office 365
Location: Switzerland

Re: TM1 10.2.2 FP6

Post by lotsaram »

AmbPin wrote:Was the function you are thinking of "DisableMTQViewConstruct"
Yes that sounds right. Although obviously nothing to do with temp views.
Please place all requests for help in a public thread. I will not answer PMs requesting assistance.
Post Reply