Saposcol communicates with the "outer world" using the shared memory and reader-writer mechanism. To avoid situations when part of the data
red from the shared memory is from one collection interval and second one from another, there is exclusive lock for writing and for reading.
If there are a lot of readers, it may time-to-time happen that saposcol is not able to lock memory for writing.
This message alone does not mean that you have some problems, this is just a problem that may happen in this case of communication and data exchange within one shared memory segment. But if you for example would have problems with failing calls to retireve data from ST06N transaction
#1692551 - Wrong log entry classification
This warninig in dev_coll was detected only under HPUX.
If this warnings appear approximately only once per hour, then according to the note 1692551 these warnings can be ignored.
There are several improvements in shared memory locking, shared memory communication mechanisms and time-consuming data
generation. It should in some cases improve the situation where several tools are attaching to same shared memory segment, where
saposcol acts as a writer and every CCMS agent (standalone sapccm4x/sapccmsr as well as integrated in sapstartsrv) as reader has
exclusive access.
The line "Collection Interval" should contain 60s value. If there is less, then please add the following line in the host_profile with which saposcol is running:
hostexec/oscol/arguments = -w60
The message in log could be useful for the support to analyze data unavailability problems.
