XI 3.1 Instalation stops at 55%

1. Recently, encounter one problem when installing XI 3.1, and google it, found the cause in
http://www.forumtopics.com/busobj/viewtopic.php?t=145857&postdays=0&postorder=asc&start=0

 

it is caused by the PRE_LOAD environment variable.

 

"Hi

I was planning to put an update on the message when I had something to report. Well, I finally have.

BO Support informed me that BO XI didn't work work if there were more than 64 cores on the server. That's fine, my server has two cpu's each with six cores. Each core is running 8 threads.

If you run psrinfo -p, it does indeed report two cpu's. However, psrinfo alone reports 96 cpu's (2x6x8). It seems Solaris treats every thread as a cpu.

I asked my hardware engineer to disable some of the threads at the system level. He reduced the number to 58.

I re-ran the install and it worked.

Now, having worked out what the problem is a quick google provides a further option of preventing libhoard from loading by commenting out the environment variable LD_PRELOAD="libhoard.so.1" and the subsequent export from env.sh in the bobje/setup folder.
Now, I've tried restarting the cms at this point but for me it didin't work. What I was forced to do was amend env.sh during the installation process itself. This also required changing attributes to make it writeable before the changes take affect. Doing it this way for me meant that the installation completed successfully which meant all I had to do was load up the BIAR file and bingo - a working application. I've done this in a test folder on my server - I have yet to deploy it properly in production.
What I did was built a single cpu/single core server just to be able to provide some kind off live service in the interim.
Although not a monster, the production server has much more grunt. Deployment is planned for this commong weekend (08/03/10). I'll post another update after the deployment if there's anything to add.

Garry Hewett"

 

checked the /var/adm/messages file, find the following error info.

 

Mar 29 05:35:26 SERVERNAME Server Intelligence Agent: server SERVERNAME.ReportApplicationServer stopped unexpectedly.Mar 29 05:35:26 SERVERNAME Server Intelligence Agent: server SERVERNAME.ReportApplicationServer failed to launch. It is possible the working directory or file path is invalid. /opt/PlexView/BOE/bobje/enterprise120/solaris_sparc/ras//boe_crystalrasd: not foundMar 29 05:35:28 SERVERNAME boe_crprocd[2423]: [ID 477990 user.error] An error occurred while creating a subprocess in the processing server. (RCIRAS0602)Mar 29 05:35:57 SERVERNAME last message repeated 3 timesMar 29 05:36:07 SERVERNAME boe_crprocd[2423]: [ID 477990 user.error] An error occurred while creating a subprocess in the processing server. (RCIRAS0602)Mar 29 05:36:37 SERVERNAME last message repeated 3 timesMar 29 05:36:41 SERVERNAME Server Intelligence Agent: server SERVERNAME.DesktopIntelligenceJobServer stopped unexpectedly.

 

2. what is hoard?

 

The Hoard memory allocator is a fast, scalable, and memory-efficient memory allocator. It runs on a variety of platforms, including Linux, Solaris, and Windows. Hoard is a drop-in replacement for malloc() that can dramatically improve application performance, especially for multithreaded programs running on multiprocessors. No change to your source is necessary. Just link it in or set just one environment variable

 

see http://www.hoard.org/