Discussion:
PowerHA for IBMi
(too old to reply)
Graap, Kenneth
2014-10-21 15:08:53 UTC
Permalink
Raw Message
I would be interested to know if anyone is using PowerHA for IBMi along with Live Partition Mobility.

Specifically how one might use these processes to perform monthly maintenance, such as loading a Technology Refresh, without interrupting service to Telnet users.

My "incomplete" understanding of these processes seems to indicate that if I want to do an IPL to apply PTF's followed by running a SAVSYS backup, I'd still have to take the system from my telnet connected application users. Is this true? If so, how can I ever have a truly 7x24 IBMi environment?

Also, using PowerHA, External Disk (V840 SAN), and VIOS, how could I automatically SWAP to my HA system in the event I lose my primary system unexpectedly (i.e. power outage or system crash)?

Kenneth
Kenneth E. Graap
NW Natural
System Administrator for IBM Power Systems
503.226.4211 x5537
http://www.linkedin.com/in/kennethgraap
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
P***@public.gmane.org
2014-10-21 17:06:25 UTC
Permalink
Raw Message
You can use Live Partition Mobility to move a running partition between systems to remove outages caused by planned hardware (compute or storage) maintenance. But if you need an OS upgrade/PTFs etc, the OS must be restarted. You know the frequency of those outages in your environment better than I do. For those outages, you can use PowerHA to reduce the downtime (my customer experiences is 15-20 minutes for a switch) but it won't zero the outage needed to move the disk assignment. For failures, PowerHA and/or HMC Remote restart can be used to automate recovery for less hands on.


Patrick Bingham | Principal, Converged Infrastructure Architect
Office: 402-965-2381 | Mobile: 402-212-2944 patrick.bingham-***@public.gmane.org
Sirius Computer Solutions | www.siriuscom.com
14301 FNB Parkway, Suite 400, Omaha, NE 68154







-----Original Message-----
From: Graap, Kenneth [mailto:Kenneth.Graap-EwKdArkbKgQS+***@public.gmane.org]
Sent: Tuesday, October 21, 2014 10:09 AM
To: Midrange Systems Technical Discussion (midrange-l-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org)
Subject: PowerHA for IBMi
Importance: High

I would be interested to know if anyone is using PowerHA for IBMi along with Live Partition Mobility.

Specifically how one might use these processes to perform monthly maintenance, such as loading a Technology Refresh, without interrupting service to Telnet users.

My "incomplete" understanding of these processes seems to indicate that if I want to do an IPL to apply PTF's followed by running a SAVSYS backup, I'd still have to take the system from my telnet connected application users. Is this true? If so, how can I ever have a truly 7x24 IBMi environment?

Also, using PowerHA, External Disk (V840 SAN), and VIOS, how could I automatically SWAP to my HA system in the event I lose my primary system unexpectedly (i.e. power outage or system crash)?

Kenneth
Kenneth E. Graap
NW Natural
System Administrator for IBM Power Systems
503.226.4211 x5537
http://www.linkedin.com/in/kennethgraap



--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org Before posting, please take a moment to review the archives at http://archive.midrange.com/midrange-l.

This message (including any attachments) is intended only for the use of the individual or entity to which it is addressed and may contain information that is non-public, proprietary, privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any use, dissemination, distribution, or copying of this communication is strictly prohibited. This message may be viewed by parties at Sirius Computer Solutions other than those named in the message header. This message does not contain an official representation of Sirius Computer Solutions. If you have received this communication in error, notify Sirius Computer Solutions immediately and (i) destroy this message if a facsimile or (ii) delete this message immediately
if this is an electronic communication. Thank you.

Sirius Computer Solutions<http://www.siriuscom.com>
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
Graap, Kenneth
2014-10-21 17:17:26 UTC
Permalink
Raw Message
Post by P***@public.gmane.org
But if you need an OS upgrade/PTFs etc, the OS must be restarted.
For those outages, you can use PowerHA to reduce the downtime (my
customer experiences is 15-20 minutes for a switch) but it won't
zero the outage needed to move the disk assignment.
Patrick, That's what I thought... That is better than a couple of hours though !
Post by P***@public.gmane.org
For failures, PowerHA and/or HMC Remote restart can be used to
automate recovery for less hands on.
So, is there ANY solution on the market that can perform a HOT SWAP of IBMi in the event of a crash?

Reply or Forwarded mail from: Kenneth E Graap
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
Roberto José Etcheverry Romero
2014-10-21 17:58:01 UTC
Permalink
Raw Message
Kenneth,
I'm not familiar with HA implementations but i believe that those 15-20
minutes are the same whether it's an intentional or accidental fail-over...
Best Regards,

Roberto
Post by Graap, Kenneth
Post by P***@public.gmane.org
But if you need an OS upgrade/PTFs etc, the OS must be restarted.
For those outages, you can use PowerHA to reduce the downtime (my
customer experiences is 15-20 minutes for a switch) but it won't
zero the outage needed to move the disk assignment.
Patrick, That's what I thought... That is better than a couple of hours though !
Post by P***@public.gmane.org
For failures, PowerHA and/or HMC Remote restart can be used to
automate recovery for less hands on.
So, is there ANY solution on the market that can perform a HOT SWAP of
IBMi in the event of a crash?
Reply or Forwarded mail from: Kenneth E Graap
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
P***@public.gmane.org
2014-10-21 18:03:48 UTC
Permalink
Raw Message
For crashes, I think most of the HA vendors would say they can do it. For IBM's solutions, I know of the below but most of our customers are hesitant to let the system bring up a target without determining the cause of the failure. I think that will change as more and more go to a SAN/hardware solution.

HMC Remote Restart (can be integrated with a monitoring solution)
PowerHA (heartbeat capability, via IP or HMC)

-----Original Message-----
From: Graap, Kenneth [mailto:Kenneth.Graap-EwKdArkbKgQS+***@public.gmane.org]
Sent: Tuesday, October 21, 2014 12:17 PM
To: Midrange Systems Technical Discussion
Subject: RE: PowerHA for IBMi
Post by P***@public.gmane.org
But if you need an OS upgrade/PTFs etc, the OS must be restarted.
For those outages, you can use PowerHA to reduce the downtime (my
customer experiences is 15-20 minutes for a switch) but it won't zero
the outage needed to move the disk assignment.
Patrick, That's what I thought... That is better than a couple of hours though !
Post by P***@public.gmane.org
For failures, PowerHA and/or HMC Remote restart can be used to automate
recovery for less hands on.
So, is there ANY solution on the market that can perform a HOT SWAP of IBMi in the event of a crash?

Reply or Forwarded mail from: Kenneth E Graap


--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org Before posting, please take a moment to review the archives at http://archive.midrange.com/midrange-l.

This message (including any attachments) is intended only for the use of the individual or entity to which it is addressed and may contain information that is non-public, proprietary, privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any use, dissemination, distribution, or copying of this communication is strictly prohibited. This message may be viewed by parties at Sirius Computer Solutions other than those named in the message header. This message does not contain an official representation of Sirius Computer Solutions. If you have received this communication in error, notify Sirius Computer Solutions immediately and (i) destroy this message if a facsimile or (ii) delete this message immediately
if this is an electronic communication. Thank you.

Sirius Computer Solutions<http://www.siriuscom.com>
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
Nathan Andelin
2014-10-21 18:11:34 UTC
Permalink
Raw Message
Post by Graap, Kenneth
So, is there ANY solution on the market that can perform a HOT SWAP of
IBMi in the event of a crash?
What do you mean by a crash? A catastrophic event that destroys a data
center in an instant? It appears to me that 99% of "crashes" are disk drive
failures, 99% of which may be mitigated by disk mirroring.

So 99.9% of outages can be planned hot-swaps.

Some folks advocate moving applications off centralized servers to server
farms managed by load balancers, so that a complete failure of a single
application server doesn't materially affect overall data center
operations. But an entire server farm can be taken down by a failure of a
single database server.

Nathan.
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
DrFranken
2014-10-21 19:36:15 UTC
Permalink
Raw Message
There are indeed a million ways to muck this up. I know well of a
company in SFO that has two MASSIVE stacks of linux boxes supporting
MySQL in 26 shards. 100% SSD, millions of transactions per hour. The two
stacks mirror each other. Never supposed to be down they say. Awesome.
Then somebody does a dumb thing and they switch to the backup server.
Then they delete a table on what was thought now to be the backup server
BUT somehow that delete replicates to current production. No matter it
was just the customer file. (SSDs delete real fast.)

There was no hardware failure or data center failure whatever. Just
people reacting incorrectly to an oopsie.

To err is human, to really muck it up takes a computer.

- Larry "DrFranken" Bolhuis

www.frankeni.com
www.iDevCloud.com
www.iInTheCloud.com
Post by Nathan Andelin
Post by Graap, Kenneth
So, is there ANY solution on the market that can perform a HOT SWAP of
IBMi in the event of a crash?
What do you mean by a crash? A catastrophic event that destroys a data
center in an instant? It appears to me that 99% of "crashes" are disk drive
failures, 99% of which may be mitigated by disk mirroring.
So 99.9% of outages can be planned hot-swaps.
Some folks advocate moving applications off centralized servers to server
farms managed by load balancers, so that a complete failure of a single
application server doesn't materially affect overall data center
operations. But an entire server farm can be taken down by a failure of a
single database server.
Nathan.
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
Graap, Kenneth
2014-10-21 19:45:01 UTC
Permalink
Raw Message
What do you mean by a crash? A catastrophic event that destroys a data center
in an instant? It appears to me that 99% of "crashes" are disk drive failures,
99% of which may be mitigated by disk mirroring.
Nathan, By a CRASH I mean the sudden termination of all processes on the system. For example, pulling BOTH power cables from the CEC while the system is running (An IBM CE actually did this to us once!).

The system issuing SRC B6000103, forcing a Main Storage Dump and initiating an Abnormal IPL (This is what happened to us 3 weeks ago!)

These are kind of system crashes I'm referring to.



Reply or Forwarded mail from: Kenneth E Graap
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
Nathan Andelin
2014-10-21 20:17:52 UTC
Permalink
Raw Message
Nathan, By a CRASH I mean the sudden termination of all processes on the
Post by Graap, Kenneth
system. For example, pulling BOTH power cables from the CEC while the
system is running (An IBM CE actually did this to us once!).
Thanks for the point of reference. I'm not aware of any products that could
provide a complete, automated, and error-free hot-swap in that type of
crash.

I think experts might suggest that synchronous mirroring provided by
PowerHA or remote journaling would provide a near error-free recovery on a
backup server in such an event. But I'm not aware of any product that might
provide an instantaneous hot-swap.

Nathan.
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request-Zwy7GipZuJhWk0Htik3J/***@public.gmane.org
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.
Loading...