The following problem has been reported against ptx/SESMON V1.2.0
If a fabric device is renamed while sesmond (the ptx/SESMON monitoring daemon) is running, sesmond does not correctly handle the name change. The result is odd behavior for single-ported enclosures, and odd error messages with dual- ported enclosures.
On single ported systems when a fabric is renamed, sesmond behaves as if an enclosure were removed and a new enclosure added- it logs messages as shown below. A side-effect of this behavior is that error states are not saved - that is, existing previosly reported faults are reported again.
On a dual-ported system, when one fabric is renamed, sesmond mistakes the renaming for the addition of a third port, and any log messages include both the old and the new names of that fabric. If both fabrics are renamed within one polling period, sesmond behaves in the same way as when a single-ported enclosure has its fabric renamed- as if the DAE enclosure has gone away and a new enclosure has been added.
Workaround: Stop and restart sesmond by invoking the appropriate shell scripts: sh /etc/init.d/sesmon stop followed by sh /etc/init.d/sesmon start.