RE: Another disk cache flush issue question...

One of my customers went to replace the controller cards as the result of
the engineering change and it lost the load source drive. After some
serious calls with IBM they put the OLD cards back in and everything worked.
This experience is similar to yours although you were able to get past it.

I think there is a problem with the new cards they are shipping


--
Jim Oberholtzer
Agile Technology Architects

-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of
DrFranken
Sent: Thursday, June 08, 2017 10:03 AM
To: Midrange Systems Technical Discussion
Subject: Re: Another disk cache flush issue question...

I think after I got all the guest partitions shut down I would vary off all
the NWSDs on the HOST. Then do ENDSYS there to bring it to restricted state.
Finally do CHGASPACT *FRCWRT on each ASP there.

Once that is done, and it might take 'some minutes', then do pwrdwnsys
*CNTRLD. Considering that you're in restricted state that should get
everything to disk.

I BELIEVE (foggy memory here) that if the flushing to disk takes too long
the system shuts down leaving data in cache. So something last time was
telling the system there was data in them there caches and not to use the
disks or damage would occur.

When you started back up and shut back down then you had enough time to
flush the caches on the second attempt so it then let you swap them.

- Larry "DrFranken" Bolhuis

www.Frankeni.com
www.iDevCloud.com - Personal Development IBM i timeshare service.
www.iInTheCloud.com - Commercial IBM i Cloud Hosting.

On 6/8/2017 10:32 AM, Rob Berendt wrote:

Power 8. IBM i hosting IBM i (and AIX). All IBM i partitions are
running IBM i 7.3.

We're replacing 57CE cards with updated 57CE cards due to an issue.

We did this on one 8286-41A and had no issue.
Quiesce all applications.
ENDWTR *ALL
ENDHOSTSVR *ALL
ENDTCPSVR *MGTC
ENDTCPSVR *CIMOM
ENDTCP
ENDSBS *ALL *CNTRLD DELAY(120)
DSPMSG QSYSOPR
Wait until message 'System ended to restricted condition' appears.
PWRDWNSYS *CNTRLD DELAY(120) RESTART(*NO) CONFIRM(*NO) Watch HMC until
it says the LPAR is shutdown.
Execute the shutdown -force command on the VIOS lpars.
Shutdown the rack using the HMC.
Replace the cards.

On our other 8286-41A we modified this slightly.
...
Wait until message 'System ended to restricted condition' appears.
PWRDWNSYS *CNTRLD DELAY(120) RESTART(*NO) CONFIRM(*NO) ...

In between those two lines we did a RCLSTG and a full system save.
On that machine when we brought it up it would not recognize the disk
drives. So we powered it all off again. Put the old cards back into
their respective slots. Powered up the system. Powered up the
hosting lpar manually. Went into DST and used that to power off
(F10). Powered off vios and the rack again. Replaced the cards with the

new ones.

Started it all back up and it worked fine. Same IBM CE.

He'll be doing another machine on Saturday. Short of hanging rosaries
on it and sprinkling it with holy water is there some way to ensure
that cache is totally flushed? Got the same IBM CE doing it. Me,
however, will be on my way out of state on a Scout trip. A coworker
will be handling this.
I am willing to document some additional steps he may need.


Rob Berendt

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx To subscribe, unsubscribe,
or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take a
moment to review the archives at http://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxx for any subscription related questions.

Help support midrange.com by shopping at amazon.com with our affiliate link:
http://amzn.to/2dEadiD