When a fault or other annunciated condition occurs, there are multiple sources of information to be evaluated prior to taking action. The evidence left by the condition is a function of the type of fault and whether the CSLS is simplex or redundant.
A fatal error in a simplex CSLS generally results in a process shutdown because output LS CHARMs of the CSLS are de-energized. COMM alarm occurs immediately since the unit goes through reset. After (if ) the failed CSLS is able to boot up the COMM alarm clears and PAST_ERROR is reported to show the transient fatal error. If the error persists (repeats within 1 minute from finishing boot up) the CSLS resets again and boots up in reduced mode posting FAILED alarm.
A non-fatal error in a simplex CSLS does not impact the process. MAINT or ADVISE alarm(s) are posted depending on the condition(s).
A fatal error on a single CSLS unit of a redundant CSLS does not impact the process because the other CSLS unit continues to drive outputs. A fatal error on a single CSLS will result in CSLS switchover if the failed unit was operating as active at the moment of failure. MAINT alarm is posted because the partner unit with the fatal error is going through reset and is not available until finished the boot up. When (if) the failed standby CSLS unit boots up OK the MAINT alarm clears. PAST_ERROR on standby CSLS is reported to show the transient fatal error. If the error persists (repeats within 1 minute from finishing boot up) the standby CSLS resets again and boots up in reduced mode posting MAINT alarm that latches until the fault is repaired.
A fatal error on both CSLS units of a redundant CSLS back-to-back with or without a CSLS switchover (or rather a fatal error on active CSLS with a faulty or absent partner) generally results in a process shutdown because output LS CHARMs of the CSLS are de-energized. COMM alarm occurs immediately since the units go through reset. After (if) the failed CSLSs are able to boot up (at least one) the COMM alarm clears and PAST_ERROR is reported to show the transient fatal error(s). If the error persists (repeats within 1 minute from finishing boot up) on both units again the CSLS units reset again and boots up in reduced mode posting FAILED alarm that latches until the fault is repaired
A non-fatal error in a redundant CSLS results in MAINT or ADVISE alarm(s) depending on the condition(s).Non-fatal errors can occur on both active and standby units and do not cause switchovers.
Some non-fatal errors that occur on standby unit would not inhibit the CSLS switchover (PAVAIL is YES), others (like NVM error) do (PAVAIL is NO). Some non-fatal errors are not fatal as a single error and become fatal in conjunction with other errors (PRI and SEC peer errors).
The next step is determining whether the error condition is still present. Typically detected faults are persistent, that is, they are caused by a hardware failure and require that the CSLS unit be replaced. But some conditions are transient, clearing after being active briefly. In this case an alarm that was posted to signal the fault clears (goes to inactive, unacknowledged state). Diagnostic parameters do not indicate the cause of the alarm because the condition is no longer active. Event Chronicle can be used to determine which condition caused the alarm when the alarm is no longer active.