Most likely to be SAN issue and not z/OS.  You could look at the RMF Ficon
report to see if there is buffer credit starvation, for example.  This can
occur even if your competing z/OS is idle but has SAN resources allocated
in the switch.  Also check the RMF ESS Link stats which would show any
competing activity (outside of your client's z/OS).  If there is
congestion, it is your SAN and switches config that needs improvement :)

On Sun, Nov 19, 2023 at 10:00 PM Alain Benvéniste <[email protected]>
wrote:

> Hi,
>
> With a z/VM7.3 with the last RSU and a z/OS2.5 I met this week a
> performance problem for a client who restores on our site.
>
> My z/OS is required to restore the z/OS client's bootstrap.
> When done I need to change several things for him to IPL.
> It means that the dasd are shared between my z/OS and his.
> I let my z/OS active, no reason to stop it.
>
> The client IPL his z/OS and sends his restores.
> Checking DASD throughput we are at 1/3 speed that it should.
> After spending the day to search, I forced my z/OS.... then all the speed
> was there for the client.
>
> I wonder how my z/OS could impact client's restore just because I access
> the dasd, doing absolutely nothing on it.
>
> This morning I did a backup of client's data, the speed seems ok.
> Now I restore and the speed is down to 1/3...
>
> It looks like z/os brakes througput…
>
> If you have any clue : I can test.
> An issue is opened.
>
>
> Regards
> Alain
>
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to [email protected] with the message: INFO IBM-MAIN
>

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

Reply via email to