>it seems the ESTAE got control before the FRR 

wanna bet?

As the earlier posts mentioned, it is possible that an FRR will never get 
control (for example if an SDWA could not be obtained, which typically 
would happen only if the system was already in deep trouble). But an ESTAE 
will never get control before an FRR (where such a statement implies that 
both do get control).  One might argue that if an FRR was skipped because 
of inability to get an SDWA that the address space should be memterm'd 
because the recovery routine would not have gotten a chance to clean up. 
An FRR (also true for ESTAE) would also be skipped if the cross-memory 
environment that it requires cannot be established.

I'd suggest that if you think this is happening, you get a dump and 
examine the system trace, specifically for RCVY entries. A RCVY FRR entry 
is provided whether the FRR is skipped or not, with an indication in the 
data if it was skipped. 

Peter Relson
z/OS Core Technologoy Design


----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to