For release r17.2 we are still in the process of analyzing evidence so that we can confirm that this is the same problem.
If confirmed that this is the same problem then the fix that was created and delivery on r14.1.02, will be ported to serve customers on release r17.2.
If you are experiencing this behavior please open a case in support informing you about this article, and attaching to the case the following information to assist in our investigation:
Enable the following trace/processes in your environment:
pdm_trace bpvirtdb_srvr on
pdm_trace bpvirtdb_srvr size 100000
pdm_trace bpvirtdb_srvr watch sqltrans.c 2156
pdm_trace bpvirtdb_srvr attach domsrvr
pdm_trace domsrvr size 100000
Traces should be written automatically due to the watch setting.
Monitor the environment for at least one day. As soon as the slow-process starts please write down the time so that we can be more assertive in our analysis of traces.
To turn off trace, including attached processes, run:
pdm_trace bpvirtdb_srvr off
Collect pdm_trace for domsrvr and bpvirtdb_srvr and stdlog files and attach to the case open in the stand.