EU_CLONE_MIG_DT_PRP running since 7 hours -

[Former Member]

Hello Experts ,

We are migrating our ECC system from Oracle to HANA DB using DMO - but the mentioned phase 57.6 % done (in phase: MAIN_SHDIMP/SUBMOD_MIG_UPTIME_PRP/EU_CLONE_MIG_DT_PRP)  is running since 7 hours with no update in the logs.

Last log updated was EUMIGRATEDTPRP and there are no errors written in it.

Few lines from the log are :

3 ETQ123 20150113065919 (2): PID 9164 exited with status 0 (time 0.000 real)

1 ETQ399 Got 18 split conditions for table 'PCL4' in 'h:\DMO_E\SUM\SUM\abap\migrate_dt\MIGRATE_DT_PRP0003_SPLIT.WHR' - adding to cache.

3 ETQ123 20150113072306 (1): PID 3656 exited with status 0 (time 0.000 real)

1 ETQ399 Got 4 split conditions for table 'PPOIX' in 'h:\DMO_E\SUM\SUM\abap\migrate_dt\MIGRATE_DT_PRP0011_SPLIT.WHR' - adding to cache.

Our environment is as followed :

• Source ECC6 Ehp5 702 on Oracle.
• Target  ECC6 Ehp7 740 on HANA.
• SUM SP12 PL02

What checks i can perform to ensure if the phase is running properly ?

What can i do to speed up the process.?

My CPU usage is 15% only and CPU usage is 5%.

Thanks

Dev

[Boris_Rubarth]

Hi Dev,

indeed this phase may take long, depending also on the performance of the source database.

You should check whether there are still processes running, esp. SAPuptool.

The number of R3load uptime processes is relevant for this phase. It can be adapted via browser, see section

4.2.2 "Changing processing parameters during runtime" in the DMO documentation

A.3.2 "Configuring Parallel Processes during Runtime" in the SUM guide

[updated on July 12th 2015].

Best regards, Boris

Message was edited by: Boris Rubarth, as the explanation is now part of the SUM guide.

[Former Member]

Hello Boris ,

After wait of 8 hours the phase is over - i had allocated 16 r3trans/r3load which i assume are sufficient as i have 18 core CI machine.May be i could have defined more. ?

We have 1.5 TB of data - which might have increased the time , SAPuptool was running but it wasnt consuming much of memory so i was worried if its running fine or not.I would expect some logs so that we are aware about the situation.

Thanks

Dev

[RolandKramer]

See the Blog - DMO: optimizing system downtime ...

Of course you can define between 8-10 processes per CPU core on a UNIX based server.

Best Regards Roland

[Former Member]

Thanks Roland - we increased our r3trans/r3loads significantly but interestingly time taken by this phase was same and yes- no logs:(

Regards

Dev

[CharlesChen]

Hi Dev,

Have you checked whether any disk space or tablespace issue there in the source system?

Best regards,

Charles.