Paul:
I've seen and worked with several of the non-IBM vendors in HA environments.
Caveat Emptor. What they show you and promise is the time it takes for the
switch to actually happen, as opposed to the entire time it takes to make
the switch including application quiesce, and all the other things that need
to be done to fail over a system, assuming a planned failover. I'm happy
when the total process is about an hour. Promising anything less is
boarding in falsity in my view.
PowerHA can actually switch in a couple of minutes. That does not mean your
applications will be up and running. Same as all the other guys.
Application replication has been around a long time and except for one
vendor I'm aware of, they all rely on remote journals. That's OK, it works.
But you have to manage the journals, be sure everything is in synch and so
on. They will tell you it takes minutes a day, true except when it does
not, which in my experience is the majority of the time.
When we quote management of systems, we have a base for that system then add
on for replicated systems. We add $0 for PHA, why, it does not take our
technicians that much longer to manage it. All others have an upcharge that
depending on the HA vendor can get as high as 30% of the base.
--
Jim Oberholtzer
Agile Technology Architects
-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of
Steinmetz, Paul
Sent: Thursday, March 29, 2018 2:37 PM
To: 'Midrange Systems Technical Discussion' <midrange-l@xxxxxxxxxxxx>
Subject: RE: What is the difference between Flash storage and a Flash
Storage system like the V9000/V7000
Paul,
I'm a bit shocked at your IPL times.
My current P7-8205-E6C takes 2 minutes to shutdown, 5 minutes to IPL.
I'm also shocked at your role swap time, 2 hours.
I've had several demos/webexs with various HA vendors, they are all claiming
less than 5 minutes role swap time.
I'm still researching the various HA/DR solutions for possible future P9
config, if corporate commits.
The P9 is a definite.
HA/DR, we'll see.
We have a 2nd data center, connected via fiber.
Paul
-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of
Musselman, Paul
Sent: Thursday, March 29, 2018 3:13 PM
To: Midrange Systems Technical Discussion
Subject: RE: What is the difference between Flash storage and a Flash
Storage system like the V9000/V7000
The question becomes: How long can the company function without the
computer? 1 day? 6 hours? 2 hours (for an IPL of the other box)? 2
minutes?
Then you attach a price tag to each answer, and see when they say 'ouch!'
It's an insurance policy.
We have a DR system using dueling iSeries/V9000 configurations. The V9000s
keep each other in sync (using Consistency Groups because "Metro Sync" ends
up as 'synchronous' instead of 'async' when the buffers fill up). No iASPs.
So the copy is a system that needs an IPL to be usable. That's about 2
hours (including futzing around, wringing hands), but that's acceptable to
the company.
We can make a flash copy of the DR copy and IPL that for auditing purposes--
showing that the live system lives on the DR box and is usable.
iASP is nice, but where does the iASP live? If it's in the data center and
the data center goes out, or the iASP dies, there goes your HA solution.
You still need an off-site copy, either on tape or in a 2nd box.
Paul E Musselman
PaulMmn@xxxxxxxxxxxxx
-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Jim
Oberholtzer
Sent: Thursday, March 29, 2018 2:46 PM
To: 'Midrange Systems Technical Discussion' <midrange-l@xxxxxxxxxxxx>
Subject: RE: What is the difference between Flash storage and a Flash
Storage system like the V9000/V7000
We are discussing the difference between and HA system and DR system, the
difference being simple. The HA box uses iASP and can do a switch in
minutes. The DR solution replicates the entire partition (the target
partition is powered off) and in the event of an unplanned switch has to IPL
and recover from an unplanned outage.
A planned switch would be to initiate the switch (use the IBM Rochester
services toolkit to make it one command) and the local system powers off,
then starts the remote box, and reverses the flow of data. No rebuild.
ALL systems should be fully journaled and that minimizes the Database
Consistency since the system simply uses the journals to get everything up.
Done properly an unplanned outage can be recovered from with relative ease
and minimal downtime.
Since the SAN is doing the replication that moves that workload off of the
Power Systems box saving quite a bit there.
--
Jim Oberholtzer
Agile Technology Architects
--
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:
https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take a
moment to review the archives at
https://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
--
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:
https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take a
moment to review the archives at
https://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
midrange.com
