On 6 June 2017 at 17:38, Jesse 1 Robinson <[email protected]> wrote:
> That APAR may look enticing, but it's pretty unlikely. Some data/time field 
> in Unicode turned all zero in late 2015. Anyone who IPLed after
> that moment in time would get Unicode initialization failure. I doubt that OP 
> has avoided IPL all this time unless it's a very infrequently used
> LPAR.

I don't think the entirety of Unicode services fails to initialize.
Rather, callers of the conversion information service CUNLINFO will
fail. Those who just use the conversion services will work OK until
z/OS 2.1. The symptoms otherwise match perfectly (rc 08/0B). Is it
possible that CSSMTP was updated some time before the IPL, and is now
using CUNLINFO where it previously did not?

The quick and definitive test is in the APAR: look at the (former)
timestamp field in the UCCB. Easily done with IPCS or TSO TEST or
ISPF, e.g. in ISPF type DDLIST, then br 10.?+220?+3C?+10

The first 8 bytes should be 0's. If they look like a TOD clock from
around the time you IPL'd, then this is your problem. Unexplained is
why it surfaced only now.

Tony H.

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

Reply via email to