In my opinion you may run into some interesting issues with timestamped 
stuff in the CFs and whatnot if the times start coming in a slightly 
unpredictable sequence.  I don't think I'd take that risk. 

When we migrated from ETR to mixed to all STP CTN, the key was two 
machines were ETR and had the STP feature, we had one machine that was ETR 
only with no STP but it was getting fed by the two ETR/STP machines.  Note 
that we had two machines as time sources, I'm also not totally sure I'd go 
with one machine connected to the 9037 and rely on that. 

I do not believe that anything z196 and up can do ETR at all.  I'm pretty 
sure I read that in the z196 Tech Guide redbook somewhere.  (Page 14 - STP 
section)

Thomas Ambros
Operating Systems and Connectivity Engineering
518-436-6433





From:   "Staller, Allan" <allan.stal...@kbmg.com>
To:     IBM-MAIN@LISTSERV.UA.EDU
Date:   02/07/2014 09:03
Subject:        Re: Sysplex Common Time Source
Sent by:        IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU>



ISTR that all time references in a SYSPLEX must be within a certain 
tolerance (I have no recollection of the actual value). 
If the local time source exceeded this tolerance, the ETR (STP/9037) would 
guide them to the same value, or failing that , remove the system from the 
plex.

The act of syncing the STP and the 9037 to the same time ETR might/might 
not provide sufficient accuracy.

HTH,

<snip>
>That was what we originally though too, but our local IBM support 
>person told us we couldn't. Thinking further about configuration 
>activities to migrate to mixed CTN mode, I'm not seeing how on the non 
>STP capable processor we're going to be able to set the name of the 
>mixed CTN, since it isn't going to have the STP tab for that CEC,

I'm not sure that that matters. IIRC, the requirement is that all systems 
in the sysplex get their time from the same source. As long as the z10 
gets it from the timers, and there's another CEC in the CTN getting it 
from the same timers and acting as the primary time server for the CTN, I 
believe that would work. 

z10 <- timer
zEC12 <- timer
 | Primary time server
\/ STP
other STP CECs

Before STP, there was no STP tab on the HMC and the CECs were able to 
participate in the sysplex by virtue of using the same timer. 

I *think* the migration would be to make the time source the timer 
connected to one zEC12, make that the PTS, then add the z10 pointing to 
the timer.

But I haven't read the fine manuals for some time, and it seems like the 
PTS becomes a sinle point of failure, unless at least one other machine in 
the CTS has a timer connection too. The point about the mixed mode CTN 
being envisioned as a fall-back situation, not an expected long-term 
situation is also good one. And if IBM is telling you "no", it's hard to 
argue that you'd want to try to do it. 
Scott
</snip>


----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN



This communication may contain privileged and/or confidential information. It 
is intended solely for the use of the addressee. If you are not the intended 
recipient, you are strictly prohibited from disclosing, copying, distributing 
or using any of this information. If you received this communication in error, 
please contact the sender immediately and destroy the material in its entirety, 
whether electronic or hard copy. This communication may contain nonpublic 
personal information about consumers subject to the restrictions of the 
Gramm-Leach-Bliley Act. You may not directly or indirectly reuse or redisclose 
such information for any purpose other than to provide the services for which 
you are receiving the information.

127 Public Square, Cleveland, OH 44114
If you prefer not to receive future e-mail offers for products or services from 
Key 
send an e-mail to mailto:dnereque...@key.com with 'No Promotional E-mails' in 
the 
SUBJECT line.

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

Reply via email to