Very low import performance of CR Content Update

[thomas_windischmann]

Hi,

I want to upgrade the CR Content for SLD. The import of the CR Content Deltas runs serveral hours. Have anyone the same experiences or suggestions to make the import faster?

The actual situation is

J2EE Server: SAP J2EE Engine/7.00 PatchLevel

SLD Version: 7.0009.20060719102106.0000

Model Version (sld/active): 1.6.9

SAP CR Content Version (sld/active): SAP_CR 2.1 (produced 05/09/2005)

Thanks in advance,

Thomas

[thomas_windischmann]

Hi Wayne,

You are right. In the meantime the jobs completed successfully, some in several hours. Because this task isn't one that you have to do very often it is tolerable.

Thanks for your acknowledgement.

Thomas

[Former Member]

Hi Thomas,

I have experienced some very long runtimes when updating the CR content, in some cases several hours. However, they always complete successfully. I've looked in to speeding this up but didn't find any way to do it.

Are your jobs completing successfully? If so, you may have to put up with long runtimes. Fortunately this task isn't one that you have to do very often so it may be tolerable.

WL

[thomas_windischmann]

Michael,

sorry for the delay, in the meantime there has been an other problem for troubleshooting.

The output of prstat -s cpu is

PID USERNAME SIZE RSS STATE PRI NICE TIME CPU PROCESS/NLWP

461 bwtadm 1585M 1319M sleep 39 0 2:28:14 1.4% jlaunch/188

2984 orabwt 4728M 4654M sleep 1 0 0:34:25 0.4% oracle/11

17261 root 1607M 1562M sleep 59 0 367:14:17 0.0% esd/4

33 bwtadm 4739M 4653M sleep 1 0 0:04:01 0.0% oracle/15

9638 bwtadm 9080K 3680K sleep 1 0 200:15:18 0.0% saposcol/1

277 bwtadm 74M 69M sleep 1 0 0:02:12 0.0% enserver/6

460 bwtadm 413M 292M sleep 41 0 0:04:39 0.0% jlaunch/116

48 bwtadm 4725M 4650M sleep 1 0 0:02:34 0.0% oracle/1

11060 bwtadm 1800K 1720K cpu0 59 0 0:00:00 0.0% prstat/1

42 bwtadm 4726M 4651M sleep 1 0 0:02:08 0.0% oracle/1

7155 noaccess 182M 76M sleep 1 0 23:24:43 0.0% java/55

24 bwtadm 4734M 4658M sleep 1 0 0:00:34 0.0% oracle/258

18 bwtadm 4726M 4651M sleep 1 0 0:02:11 0.0% oracle/1

26 bwtadm 4732M 4657M sleep 27 0 0:00:35 0.0% oracle/258

3745 root 156M 11M sleep 1 0 17:36:51 0.0% vxsmf.bin/8

Total: 160 processes, 2051 lwps, load averages: 0.78, 0.78, 0.63

The SLD is installed on a server (16 GB RAM) with Sun Microsystems Inc. 1.4.2_15 on SunOS 5.10, on which a SAP BI 7.0

is running, too. DB is Oracle 10.2. In the meantime we have installed SAP_CR 6.1 (produced 04/29/2010).

Regards,

Thomas

[Former Member]

Do you see many garbage collections? Or full gc's?

Kind regards,

Mark

[Former Member]

Which process is consuming the most cpu during the load? Let's have a look what is happening on the server. What platform/os/db do you have?

Cheers Michael