> -----Original Message-----
> From: IBM Mainframe Discussion List
> [mailto:[EMAIL PROTECTED] On Behalf Of Paul Gilmartin
> Sent: Thursday, June 16, 2005 3:49 PM
> To: [email protected]
> Subject: Re: EIBTASKN - maximum value 999999 or 99999 ?
>
>
> In a recent note, John Bohumil said:
>
> > Date: Thu, 16 Jun 2005 15:06:29 -0500
> >
> > So, this isn't really any better than just using the date
> and time. EIBTASKN is
> > probably the safest because it's going to be unique for
> tasks executing nearly
> > simultaneously. I learned this the hard way when I was
> using date/time to devel
> > op a TS Queue only to find out when two transactions occur
> in the same 1/10 sec
> > they would generate duplicate TSQ names. I switched to the
> EIBTASKN and all was
> > well.
> >
> What ever became of the guarantee that STCK would return unique values
> (subject to clock wrap)?
The above was about the result from an EXEC CICS ABSTIME. It truncates
the STCK to the nearest 1/10 of a second, then extends that to
milliseconds by appending zeros.
>From the current PoPS (section 4.6.1.4)
<quote>
Two executions of STORE CLOCK or STORE CLOCK EXTENDED, possibly
on
different CPUs in the same configuration, always store different values
of
the clock if the clock is running. If the clock is stopped, zeros
are
stored in the clock value, bits 8-111 of the storage operand, in
positions
to the right of the rightmost bit position that is incremented when
the
clock is running. The programmable field continues to be stored even
when
the clock is stopped.
</quote>
>
> (Does STCKE make similar assurance?)
See above.
>
> (Is there any similar assurance across a sysplex?)
A sysplex requires either an ETR (ETRMODE YES in CLOCK00), or if all
members are on a single CEC, the SIMETRID parameter in the CLOCKnn
member of PARMLIB to be identical on all members. From the PoPS on an
External Timer Reference in section 1.3
<quote>
The external-time-reference facility provides a means to initiate and
maintain the synchronization of TOD clocks to an external time
reference (ETR). Synchronization tolerance of a few microseconds can
be achieved, and the effect of leap seconds is taken into account.
The facility consists of an ETR sending unit (Sysplex Timer*), which
may be duplexed, two or more ETR receiving units, and optical-fiber
cables. The cables are used to connect the ETR sending unit, which is
an external device, to ETR receiving units of the configuration. CPU
instructions are provided for setting the TOD clock to the value
supplied by the ETR sending unit.
- The ETR automatic-propagation-delay-adjustment function adjusts
the time signals from the ETR to the attached processors to
compensate for the propagation delay on the cables to the
processors, thus allowing the cables to be of different lengths.
(September, 1991)
- The ETR external-time-source function synchronizes the ETR to a
time signal received from a remote location by means of a
telephone or radio. (September, 1991)
</quote>
and, from "z/OS: MVS Setting Up a Sysplex"
<quote>
When the sysplex consists of multiple MVS systems running on two or
more processors, MVS requires that the processors be connected to the
same Sysplex Timer. MVS uses the Sysplex Timer to synchronize TOD
clocks across systems.
For a multisystem sysplex defined on a single processor (under PR/SM
or VM) the SIMETRID parameter in the CLOCKxx parmlib member must
specify the simulated Sysplex Timer identifier to synchronize timings
for the MVS systems.
</quote>
So, since the STCK/STCKE is guaranteed to give a unique value on a
single system. And the ETR causes all systems to have synchronized TOD
clocks, the value of an STCK/STCKE is guaranteed to be unique withing a
sysplex. Well, I admit that I may be reaching on this. But I do remember
reading that an ETR guarantees that the result of an STCK/STCKE is
unique across all members of a Sysplex. I just cannot remember where I
read it.
>
> -- gil
> --
--
John McKown
Senior Systems Programmer
UICI Insurance Center
Information Technology
This message (including any attachments) contains confidential
information intended for a specific individual and purpose, and its'
content is protected by law. If you are not the intended recipient, you
should delete this message and are hereby notified that any disclosure,
copying, or distribution of this transmission, or taking any action
based on it, is strictly prohibited.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
