Re: BRMS flash copy -- MIDRANGE-L

That doesn't happen on SVC-derived storages (Storwize, Flashsytems), you
have to either follow the wizard and hope the code doesn't have a hiccup
(Where the code says "All's OK" but when you check you have an unused disk
and a phantom dead disk) or just replace it by CLI otherwise it'll be left
as "unused".
On the bright side, since the SAN has its own Model/Type/Serial and
interface, you won't get an IBM i tech trying to replace a VIOS disk.

On Fri, 20 May 2022 at 09:25, Jim Oberholtzer <midrangel@xxxxxxxxxxxxxxxxx>
wrote:

On the SAN, if the yellow light is blinking, simply pull the drive, stuff
the new one in and forget it. I check back in about 10 minutes and
usually the dead drive that was replaced winds up being the new hot spare.
I’ve walked customers through that process several times.

With respect to VIOS disk failures, SAN boot it. Eliminate the internal
drives completely.

Jim Oberholtzer
Agile Technology Architects

On May 20, 2022, at 7:13 AM, Rob Berendt <rob@xxxxxxxxx> wrote:

I'm hoping the people who change drives in a SAN when they fail are a

lot more skilled then the service reps IBM sends out to fix Power internal
drives for VIOS. EACH. AND. EVERY. TIME. the first thing the service rep
tries to do is get STRSST working from a VIOS command line.

I can tell this is always going to happen when the dispatcher asks me

what OS is running on the lpars that the VIOS is supporting.

There is no possible communication to avoid this situation. The people

manning the phones cannot comprehend the situation. So you just let them
send the wrong guy, he flubs it, then he calls someone on the phone who can
talk him through it.

The IBM i hosted drives were a little better although not only did I

have to pull up IBM documentation but I had to call his supervisor and tell
him the education he was given was bad and needed correction. The rep
always said to wait 0 seconds to pull the drive. Then waited until AFTER
the 18 seconds he was allowed to pull the drive expired before he pulled
the drive. We never lost data but we did have to clear several errors.
The rep agreed that the IBM doc I pulled looked legit and the supervisor
agreed with me that some reeducation was in order.

And having to speak to multiple duty managers on each call to get

anything done gets annoying.

Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 7310 Innovation Blvd, Suite 104
Ft. Wayne, IN 46818
Ship to: 7310 Innovation Blvd, Dock 9C
Ft. Wayne, IN 46818
http://www.dekko.com

-----Original Message-----
From: MIDRANGE-L <midrange-l-bounces@xxxxxxxxxxxxxxxxxx> On Behalf Of

Jim Oberholtzer

Sent: Friday, May 20, 2022 7:47 AM
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxxxxxxxx

Subject: Re: BRMS flash copy

CAUTION: This email originated from outside of the organization. Do not

click links or open attachments unless you recognize the sender and know
the content is safe.

QEDD is a good product, several of my customers use it. That said, I

always go with SAN replication where possible. As a plus, it pushes the
replication workload to the SAN from the POWER system. With the toolkit,
flash copy, nightly full system saves, etc are a breeze. The BRMS iASP
version, I just can’t see where that is a solution for all but limited
situations.

Now in Rob’s case where procedures are clear and practiced, I just don’t

see a reason to change it. With respect to upgrades, moving to SAN storage
from internal storage, that is definitely the way to go, even with needing
VIOS to virtualize everything. Once you’ve got that set up properly
management is easy, can be done while the system is active etc.. Once you
get to that type of storage, you won’t choose to go back.

Jim Oberholtzer
Agile Technology Architects

On May 20, 2022, at 6:30 AM, Rob Berendt <rob@xxxxxxxxx> wrote:

I'm trying to decide if any of these really help in our situation.
GDIHQ is production.
GDIHQ2 is CBU.
We're using Quick-EDD to replicate between these.
We back up GDIHQ quarterly with a full system save. After doing a role

switch to GDIHQ2.

We do nightly backups on GDIHQ2 with BRMS.
IBM just added the capability to BRMS to create a recovery report for

GDIHQ based on this exact scenario.

I have zero quisce operations and zero disruption of the production

machine for the nightly saves.

Back when we only had a single system the boss would not let me quiesce

to start the checkpoint for a save while active. I had to do a ragged save
while active. Of course, save while active on IFS is a sick joke.

When we do an OS upgrade we upgrade GDIHQ2 first. Let's say Wednesday

or Thursday. Then we switch over to GDIHQ2 on Saturday and upgrade GDIHQ
to the new release. This way we have zero worries about saving to a
previous release being an issue. Been through several upgrades, ptfs, etc
with no issues following this. Sunday we switch back. They're in
different cities so they have different IP addresses, etc. There is a
slight drop in 5250, etc when we do the switches. We have an automated
process which changes the DNS entries in our Windows servers.

GDIHQ2 is not a 'test'. It is CBU. We have other machines and lpars

for testing of the new releases, etc.

What we have may only sound simple to me because I'm familiar with it.
While PowerHA and/or other SAN based solutions sound simple to you

because of your familiarity with those processes.

Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 7310 Innovation Blvd, Suite 104
Ft. Wayne, IN 46818
Ship to: 7310 Innovation Blvd, Dock 9C
Ft. Wayne, IN 46818
http://www.dekko.com

-----Original Message-----
From: MIDRANGE-L <midrange-l-bounces@xxxxxxxxxxxxxxxxxx> On Behalf Of

Rob Berendt

Sent: Friday, May 20, 2022 7:12 AM
To: Midrange Systems Technical Discussion <

midrange-l@xxxxxxxxxxxxxxxxxx>

Subject: RE: BRMS flash copy

Thank you.

Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 7310 Innovation Blvd, Suite 104
Ft. Wayne, IN 46818
Ship to: 7310 Innovation Blvd, Dock 9C
Ft. Wayne, IN 46818
http://www.dekko.com

-----Original Message-----
From: MIDRANGE-L <midrange-l-bounces@xxxxxxxxxxxxxxxxxx> On Behalf Of

Rameez Raza

Sent: Thursday, May 19, 2022 4:16 PM
To: Midrange Systems Technical Discussion <

midrange-l@xxxxxxxxxxxxxxxxxx>

Subject: Re: BRMS flash copy

CAUTION: This email originated from outside of the organization. Do not

click links or open attachments unless you recognize the sender and know
the content is safe.

Dear Rob,

Flashcopy backup using PowerHA requires IBMi cluster and iASP.

BRMS flashcopy does not require PowerHA and hence can be implemented for
system having iASP or Sysbase only. You can use BRMS flashcopy to take
backup as under

On Source Side ( SYSTEMA )
To do the FlashCopy backup of SYSTEMA. You need to follow the steps
1)End SYSTEMA to Restricted State.
2)Check that System has reached to restricted state ( There are API to

do

this ). When you put the system in restricted state Main Memory of

SYSTEMA

would be flushed and all contents would be written to disk. This way you
would reduce chances damaged object on FLASHSYS ( Flash Copy System of
SYSTEMA )
3) Change the IPLA of SYSTEMA to start in restircted state
4) Set the source system to Flash Copy State ( There are commands to do
this )
5) Indicate to SYSTEMA that backup is in progress on FLASHSYS ( Though
backup has not started )
6) Now Split the association between Source Disk Array and the Flash

Disk

Array. This is called ENDPPRC link in IBM term and Split in EMC Term. A
number of checks can be incorporated to make this step more robust
7) Once the split is complete , Change the IPLA of SYSTEMA to start the
system normally
8) Start QCTL on SYSTEMA
9) Do Not Run Any backup on SYSTEMA till the time backup is running or
better say that till the time the backup completes on FLASHSYSA and

QUSRBRM

( BRMS data Library ) is restored to SYSTEMA.

Target Side ( FLASHSYS System Side)
1) IPL the FLASHSYS after the split is complete or after the ENDPPRC is
complete
2) FLASHSYS will come up in Restricted state or it will stall if it cant
find the load source disk ( if the split is failure )
3) Change the Flash state from source to flash target ( there are

commands

to do this , its easy)
4) Remove Production IP address, Remote Route ( if required ) and assign
new ip adress and new Route. New IP address means dynamically creating

line

description with correct resource name ( its easy )
5) Set Restricted state TCP/IP interface for BRMS to talk to other

systems

in BRMS network
6) Do the backup
7) Irrespective of weather backup completes normally or abnormally, save
entire QUSRBRM in to save file and FTP it to source system SYSTEMA (

to be

done as batch FTP )

Source Side ( SYSTEMA ) - Steps would be executed as part of FTP job

from

FLASHSYSA )
1) Clear QUSRBRM
2) Add to system reply list default reply of I to ignore deletion of
unsaved journal receiver
3) Restore QUSRBRM
4) Ensure the QUSRBRM is restore succesfful. If yes then indicate to
SYSTEMA that backup is now compelete on FLASHSYS.
5) End the Flash Copy on source.

On Thu, 19 May 2022, 19:34 Rob Berendt, <rob@xxxxxxxxx> wrote:

I hear talk of something called BRMS flash copy.
Is this something which still requires two lpars and/or systems or can

you

do a flash copy, and start the backup from the flash copy?
Do you have to quisce the system during the flash copy?

Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 7310 Innovation Blvd, Suite 104
Ft. Wayne, IN 46818
Ship to: 7310 Innovation Blvd, Dock 9C
Ft. Wayne, IN 46818
http://www.dekko.com

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription

related

questions.

Help support midrange.com by shopping at amazon.com with our affiliate
link: https://amazon.midrange.com

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription

related questions.

Help support midrange.com by shopping at amazon.com with our affiliate

link: https://amazon.midrange.com

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription

related questions.

Help support midrange.com by shopping at amazon.com with our affiliate

link: https://amazon.midrange.com

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription

related questions.

Help support midrange.com by shopping at amazon.com with our affiliate

link: https://amazon.midrange.com

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription

related questions.

Help support midrange.com by shopping at amazon.com with our affiliate

link: https://amazon.midrange.com

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription

related questions.

Help support midrange.com by shopping at amazon.com with our affiliate

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

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription related
questions.

Help support midrange.com by shopping at amazon.com with our affiliate
link: https://amazon.midrange.com