TM1 Forum

It appears that there is an issue with the client side release of 9.4.1 FP2.
See here and here. It would appear to be a pretty fatal issue..

Second hand advice from IBM appears to be for clients only not to upgrade to the client to FP2 and wait for FP3 to be released.

This is of course not advice from this forum or me and you should in any case make sure test everything before you upgrade...
Cheers,

Well, having upgraded to 9.4.1 FP2 and had front-end problems (embarassing TM1 problems) I was advised by IBM support to downgrade to FP1.

So, I uninstalled FP2, installed MR1 then installed FP1 ( have to do it that way). Got new error when trying to select using Subnm "cannot find macro 'evaluatearg'".

Advice from support???... known issue in FP1, upgrade to FP2.


ARGGGGGGGGGGGGGGGGGGGGGGGG.............................


So, uninstalled FP1, reinstalled MR1. Shame about anything that is fixed in FP1 and FP2...

Plus point - during the 1.5 hours my pc was tied up installing\uninstalling\re-booting\testing etc I have got rid of a lot of the rubbish on my desk!

lmao! I bet the fridge was emptier too!

So another problem with the FP2 client is when connecting across a firewall. It connects to the server, immediately disconnects and returns the message "user does not have permission to browse selected server".

Uninstalled and reinstalled FP1 and it is working.

G

There is now a Hot Fix 16 for use along with FP2 that will eliminate some of the issues until FP3 is available.

Hot Fix 16

I'm sure I have that compilation album somewhere

I've lost faith. Next implementation I'm going to reccomend 8.42 on Excel 2003

Hello all

we have a service request out with IBM (our maintenance contractors) on the performance of our new TM1 installation, 9.4.0 over Vista and Excel 2007, particularly regarding cube calculations.

we have exactly rebuilt a cube that we have in our 9.0 version of TM1 and consistently are getting the same results... less than 1 second in 9.0 compared with 13-15 seconds in 9.4 (with barely any data in the cube)

Firstly they suggest we upgrade to 9.4.1 FP1 or FP2 (see postings in this chain) but have now come back with the following:

"Upgrade both the TM1 Server and the TM1 Client installation either to TM1 9.4.1 FP2 HF16 or to TM1 9.5.0, as these are the latest and greatest (and fastest) versions of the new TM1 Server engine. After the upgrade, please rerun the performance tests."

We are not going to upgrade to 9.5 anytime soon, it doesn't fit with our upgrade policy - but regarding the HF16 option... has anyone had any experience of any problems or known issues with this version? I am particularly interested in solving the loop problem in this very thread as we are both heavy users of DBRW and SUBNM formulae

thanks in advance

Matt

Hi Matt,

I had a client that had performance issues on the web in 9.4 FP1 compared to 9.1. We put on FP2 HF16 and it improved to better that the 9.1 experience. Their problem was only with the web, so I don't know what it did for Perspectives or Excel, but they tend to run most things through web anyway, so ???

Cheers
Mal

Similar experience here. 9.4.1 FP2 with HF16 is noticeably faster than 9.1 in the Cube Viewer and Excel. However, as ever I am sure that it depends on the particular cube. Still not sure whether it is faster than 9.0 though.

One annoying issue we have is that Edit Formula insists on prefixing the cell references with the Sheet Name so instead of DBRW( A1, B6, ... you get DBRW( 'Revenue'!A1, 'Revenue'!B6...). Some of our cubes have a lot of dimensions and long sheet names which can mean that Edit Formula fails. This is fixed in 9.5.

We have also had a few unexplained crashes in 9.4.

The worst bug is the one where if you use a subset as a data source in TI, and then update a different dimension, the TI process will clear the source subset.

Another annoying difference is that it will not display cube sizes of control cubes in the properties window.

Experience with ActiveForms has been disappointing. For a view of any size they are very slow. Implementing ActiveForms also seems to mean that TM1 now recalculates all sheets as soon as you connect, which can take a while. Personally I am hoping for something better from the Cognos suite of tools.

I haven't looked at 9.5 in detail, but I would certainly advise anyone upgrading to consider 9.5 instead of 9.4.

Regards


Paul Simon

We have had to roll back to 9.4 MR1. We are waiting the release of FP3, which we thought would be out in January.

Has anyone heard anything about the intended release date of FP3, or are all these fixes being incorporated in the 9.5 rollout?

I would also be interested in hearing on any feedback on 9.5

Thanks as always for your feedback.

Armand

Armand Mizan wrote:We have had to roll back to 9.4 MR1. We are waiting the release of FP3, which we thought would be out in January.

Has anyone heard anything about the intended release date of FP3,

It's out:
http://forums.olapforums.com/viewtopic.php?f=3&t=2156

Hi,

I am running FP3 and am getting a lot of lock up problems as you may see from my two current posts.

So has anybody noticed a performance improvement moving to 9.4.1 FP3? We just upgraded to 9.4.1 FP2 from 9.0 and out clients are experiencing major performance issues when connecting excel workbooks direclty to our cubes.

I am interested if anybody has any comments...did you just upgrade your clients or did you upgrade both the server side and the client side?


thanks.

dludwig wrote:So has anybody noticed a performance improvement moving to 9.4.1 FP3? We just upgraded to 9.4.1 FP2 from 9.0 and out clients are experiencing major performance issues when connecting excel workbooks direclty to our cubes.

We had significant performance issues with 9.4 FP2, and had to roll back to 9.4 FP1.When we upgrade it will be either 9.4 FP3 or 9.5, which is in line with IBM's recommendation.
I am also trying to get input on users experience with 9.4 FP3 or even with 9.5, but am not getting much feedback.

Cheers

Armand

I did some testing and if I pull the SAME exact data out of a cube which has all of the live revenue rules the data pull takes 6 seconds (which is exactly what i would expect). When I pull that same data out of a separate cube on a separate TM1 service on a separate server which is pasted values of the cube data that has live rules on it, the excel sheet takes 65 seconds to recalculate (there are NO rules at all on any cubes on this tm1 service).

The config files are the same, the servers specs are the same, the only difference is the set of cubes without any of the rules where we have the performance issue is on a server with two other tm1 services that are used during the day. The set of cubes with the rules on it is on a server with one other tm1 service but it is just an ‘administrative’ service we have to be able to reboot the server if TM1 gets locked up where we can’t even log into it (no users ever log into this service and there are no cubes on this service).

Any ideas of what I could check for the incredible difference in performance? It is amazing to me that a cube with NO rules on it at all would take 10x as long to pull the data as compared to the cube with rules on it.

Although the dataset is the same are your views in fact identical? There can be dramatic differences in performance if an (apparently identical) view uses dynamic subsets vs. static or if large title dimensions have a selection of subset ALL rather than a small list.