TM1 Forum

Alan, Martin, Steve, Mike, Michel (did I miss anyone?)

It might be worthwhile having a 9.5.2 sticky post for a few months or else dedicating a section of the FAQ to "new features of 9.5.2" as there is a lot of new ground in terms of statistical functions and support for EP Contributor features in addition to the significant locking model changes.

9.5.2 might cause quite a bit of confusion for a while (that probably few people here will be able to answer as we're too busy supporting "legacy" versions).

Admin note: stickied on 16 Mar, unstickied on 4 May and replaced with a summary.

lotsaram wrote:Alan, Martin, Steve, Mike, Michel (did I miss anyone?)

John and Eric, but they're not around very much these days. We have two Steves, but the one above can be read as "Generic Steve" so that's OK. Matt and David are technically moderators but de facto Admins.

So that's about... OHMYGOD WE FORGOT JIM!!!!

lotsaram wrote:It might be worthwhile having a 9.5.2 sticky post for a few months or else dedicating a section of the FAQ to "new features of 9.5.2" as there is a lot of new ground in terms of statistical functions and support for EP Contributor features in addition to the significant locking model changes.

9.5.2 might cause quite a bit of confusion for a while (that probably few people here will be able to answer as we're too busy supporting "legacy" versions).

Good idea. I'll gather up some of the information that has already been posted and will shove it into here as well when I have a moment.

From the thread Migrating to TM1 v9.5.1 or TM1 v9.5.2?:
--------------------------------------

lotsaram wrote:From my understanding 9.5.2 is a MAJOR release which implements cube data versioning trees. This should virtually eliminate locking and the need in prior versions to partition high concurrency planning models into multiple data entry sub-cubes in order to prevent locking and preserve performance with lots of write-back users.

This is a major change to say the least. It should be AWESOME (once the bugs that I'm sure are there are ironed out) but my advice would be to test very well and regression test everything before deploying to a production environment

--------------------------------------

mattgoff wrote:9.5.2 changelog: https://www-304.ibm.com/support/docview ... s=swgimgmt. Very long list for an incremental/sub-minor release....

--------------------------------------

stephen waters wrote:in general, if you have an existing system you should be cautious about upgrading to a brand new software release unless there is some specific functionality you need.

With Version 9.5.2 in particular I would be especially cautious since I think the release is much more significant than the incremental naming (9.5.1 -> 9.5.2) suggests. There are major changes\enhancement to the database locking and processing which look very good but may take some time to settle down. The last time they made such significant changes was Ver 9.0 -> Ver 9.1 (admittedly pre-IBM days) and many on the forum will remember how painful that was.

One reason to go 9.5.2 provides Excel 2010 (32 bit) support (see https://www-304.ibm.com/support/docview ... 27020888#1). Several of our clients have been waiting for this for some time. Another reason is if you have immediate need of the new locking mechanism. Otherwise I suggest you upgrade to 9.5.1, applying all the hotfixes, which should give you a stable, established platform. Look at 9.5.2 in a few months time when any rough edges have been smoothed out.

From the thread How to consolidate dimension with maximum of children?:

kpk wrote:In the 9.5.2 NewFeatureGuide I have found the following:

"New functions for Rules and TurboIntegrator
The following new functions are available in TM1, version 9.5.2 for use with rules and TurboIntegrator:
● ConsolidatedMin
● ConsolidatedMax
● ConsolidatedAvg
● ConsolidatedCount
● ConsolidatedCountUnique

Note: These functions were not documented in the standard TM1 9.5.2 documentation. For complete
details, search for these functions in the latest release documentation and updates on the IBM®
Cognos® TM1® Info center."

New Features:

• Parallel interaction - A new TM1 object locking approach that increases user
throughput by lessening situations that result in write transactions that are
"blocked" by concurrent read and write transactions.
• Data reservation - A new TM1 cube level configuration that provides the ability
for users to gain "exclusive write access" to designated data slices of a cube,
preventing other users with write access updating the reserved data.
• Multi-node edit - A new TM1 Contributor capability that provides contributors and
approvers the ability to perform write back on data representing more than one
node in an approval hierarchy.
• TM1Top - Incremental capabilities to existing TM1Top that provide operational
administrators insight into sandbox and job queue server activity.
• TM1 Widget with IBM Cognos Business Insight conformance - TM1 Widgets
support a complete set of toolbar capabilities on the IBM Cognos Business Insight
toolbar. Prior releases of TM1 require the native TM1Web toolbar in the body of
the Widget.
• Added five languages - Traditional Chinese, Korean, Brazilian Portuguese, Polish,
Russian.
• Enhanced job queuing - Job queues are processed by multiple threads when used
in conjunction with parallel interaction.
• Conformance upgrade for TM1 Package Connector - The TM1 Package Connector
includes conformance for IBM Cognos Business Intelligence V10.1.0.

Just upgraded my TM1 9.5.1 test environment with Cognos BI 10.1 without any issues.

Think I am an admin, albeit I don't do much administering - those who have seen my desk will know why
When I get time to take and shake 9.5.2 I'll post some findings. My focus (as ever) will be on calculation intensive applications. If readers can recall my posting about very large models, well that's still around. It's OK though, we can fully calculate one version in two overnight runs (32 hours) and keep the server down to around 100gig while doing so - don't know what the fuss is about . Realtime, shmealtime.

Hi David,

Once I join ACG in a couple of weeks I'll also be hitting 9.5.2 with some large models. I'm currently stuck on 9.1.3 so I'll be very interested to see the difference,

Jim.

PS. Alan, how could you possibly forget me??

jim wood wrote:PS. Alan, how could you possibly forget me??

You were in the loo ("head", where you're going) when we all piled back into the van at the service station ("gas station"). Thankfully we were only 20 km ("12.4 miles", I don't envy you that one) down the motorway ("freeway") just before the 9.5.2 turnoff when Martin looked over into the back seat and said "Hey, aren't we missing someone?"

I'm looking forward to getting a 9.5.2 session in place as well, though I don't have time to do it at the moment. I'm going to be curious to see the difference in startup memory consumption between 9.0, 9.5.1 and 9.5.2 for the same model. (Not that that'll necessarily be indicative of how it runs, but it'll be interesting to see anyway.)

Alan Kirk wrote:I'm looking forward to getting a 9.5.2 session in place as well, though I don't have time to do it at the moment. I'm going to be curious to see the difference in startup memory consumption between 9.0, 9.5.1 and 9.5.2 for the same model. (Not that that'll necessarily be indicative of how it runs, but it'll be interesting to see anyway.)

Likewise (but unlike "the good old days" as a consultant it might take some time to get a test environment.) From conversations I've have with product management and engineering I'm not sure that 9.5.2 would use that much more memory than 9.5.1 on startup but once the model is in use with several dozen or several hundred users hammering away then 9.5.2 would use substantially more CPU and more likely than not would also use substantially more memory as well. It will be a wait and see whether the concept to "make better use of available CPU resources" also causes a re-think of current PVU licensing structure ...

Alan Kirk wrote:Thankfully we were only 20 km ("12.4 miles", I don't envy you that one)

makes no difference seeing as both the UK and US count distance in miles screw you Europe and the rest of the world!

Steve Vincent wrote:makes no difference seeing as both the UK and US count distance in miles screw you Europe and the rest of the world!

But the motorway marker post in the UK are in Kilometres. And remember that US miles only have 6 imperial pints to the gallon, or something.

Btw there are the 2 Steves (V and R) and I am a Stephen.

Hi

Someone mentioned that 9.5.2 supports Excel 2010. Is that the case? I couldn't see it in the release notes. I would try it but the last time I tried 2010 with TM1 9.5.1 we had to manually edit the registry to back it out.

Regards

Paul

TM1 9.5.2 Perspectives supports Microsoft Excel 2010 according to the Software Environments supported by TM1 9.5.2 ->

https://www-304.ibm.com/support/docview ... wg27020888
Cognos TM1 9.5.2 Software Environments

https://www-304.ibm.com/support/docview ... 20888#arch

lotsaram wrote: From conversations I've have with product management and engineering I'm not sure that 9.5.2 would use that much more memory than 9.5.1 on startup but once the model is in use with several dozen or several hundred users hammering away then 9.5.2 would use substantially more CPU and more likely than not would also use substantially more memory as well. It will be a wait and see whether the concept to "make better use of available CPU resources" also causes a re-think of current PVU licensing structure ...

Word from the demo the other day is that there is a 10-30% memory overhead if you're making use of the parallel writeback

PaulSimon wrote: Someone mentioned that 9.5.2 supports Excel 2010. Is that the case? I couldn't see it in the release notes. I would try it but the last time I tried 2010 with TM1 9.5.1 we had to manually edit the registry to back it out.

Yes. 2010 is supported now.

stephen waters wrote:

Steve Vincent wrote:makes no difference seeing as both the UK and US count distance in miles screw you Europe and the rest of the world!

But the motorway marker post in the UK are in Kilometres. And remember that US miles only have 6 imperial pints to the gallon, or something.

Btw there are the 2 Steves (V and R) and I am a Stephen.

I think Lotsa was just referring to the Admins; we have over 2,660 members now and it'd be hard to list them all! I'm betting that we have a bunch of Ste/v/ph/ens amongst them.

PaulSimon wrote: Someone mentioned that 9.5.2 supports Excel 2010. Is that the case? I couldn't see it in the release notes. I would try it but the last time I tried 2010 with TM1 9.5.1 we had to manually edit the registry to back it out.

As Moby 91 and Martin mentioned, it's supported but it's worth being aware of this potential issue that came through the feeds the other day:

Problem(Abstract)To start TM1 Perspectives Version 9.5.2 on Microsoft Excel 2010 fails with an error message:

Microsoft Visual Basic for Applications
Object library invalid or contain references to object definitions that could not be found

EnvironmentTM1 Perspectives Version 9.5.2 on Microsoft Excel 2010

Resolving the problemUsing a file manager like Windows Explorer go into the user specific directory "%AppData%\Microsoft\Forms\" and delete the EXD-files "RefEdit.exd" and "MSComctlLib.exd", then restart TM1 Perspectives.

I don't know how common it is, but it's not a TM1-Specific issue since .exd files are an Office thing.

Martin Ryan wrote:

PaulSimon wrote: Someone mentioned that 9.5.2 supports Excel 2010. Is that the case? I couldn't see it in the release notes. I would try it but the last time I tried 2010 with TM1 9.5.1 we had to manually edit the registry to back it out.

Yes. 2010 is supported now.

2010 on 32 bit. 64 bit wil be supported later. My guess with Cognos 10

I have successfully installed the complete TM1952 on a 32bit W7 Pro notebook.
As far as I remember I could not install the TM1web part of TM1950 and TM1951 on this notebook and were forced to use virtual machines.

kpk wrote:In the 9.5.2 NewFeatureGuide I have found the following:

"New functions for Rules and TurboIntegrator
The following new functions are available in TM1, version 9.5.2 for use with rules and TurboIntegrator:
● ConsolidatedMin
● ConsolidatedMax
● ConsolidatedAvg
● ConsolidatedCount
● ConsolidatedCountUnique
Note: These functions were not documented in the standard TM1 9.5.2 documentation. For complete
details, search for these functions in the latest release documentation and updates on the IBM®
Cognos® TM1® Info center."

Has anyone found further documentation of these new functions from some source?

Ha! I'm still awaiting access to that stupid IBM site <grumble moan winge>

mikko wrote:

kpk wrote:In the 9.5.2 NewFeatureGuide I have found the following:

"New functions for Rules and TurboIntegrator
The following new functions are available in TM1, version 9.5.2 for use with rules and TurboIntegrator:
● ConsolidatedMin
● ConsolidatedMax
● ConsolidatedAvg
● ConsolidatedCount
● ConsolidatedCountUnique
Note: These functions were not documented in the standard TM1 9.5.2 documentation. For complete
details, search for these functions in the latest release documentation and updates on the IBM®
Cognos® TM1® Info center."

Has anyone found further documentation of these new functions from some source?

I've got a service request in on it.

I have been testing the new Data Reservation functionality in 9.5.2, with the hope that a data reservation might protect the calculation cache. Sadly, I haven't found that it does. But this is very new indeed so I'm hoping others might be able to see how it could work.

Approach
a Set UseCubeVersioning=T in the CFG file
b Logon as admin
c Write and run a TI:
CubeDataReservationAcquire(
'Profit and Loss',
'admin',
0,
'|Actual||||||'
);
d Calculate the top of the cube
e Open up }StatsbyCube, wait for a minute, check that a good size calculation figure shows
f Start a new session, log on as another user
g Confirmed that the new user cannot update Actuals
h Put a value into another version
i Switch back to }StatsbyCube, wait for a minute, calculation figure now shows zero.

B&%%$r.

According to the IBM Technote 1459531

https://www-304.ibm.com/support/docview ... wg21459531
TM1 9.5.2 Cube Versioning feature renamed to Parallel Interaction


the new TM1 9.5.2 server configuration parameter "UseCubeVersioning=T" has been renamed to "ParallelInteraction=T".