Hi Michael,

I do have presentation about pros and cons for HW replication (storage based) and software replications (based on journaling) which I present on technical conferences.  These days 99% of implementation use FSR solution. Storage and network is getting faster and cheaper which makes HW DR solution more affordable and less complicated. 

From SW replication vendors, you are going to hear:

Advantages: 

• DR is accessible for test, reporting,
• DB logical errors will not be replicated
• DR node can be used for HA and offsite backups
• No changes in the application
• in depended OS instance 
• Do not replicate TEMP storage
• OS upgrade by the switch (perform OS upgrade on DR node in advance, and perform the switch). Business outage related to OS upgrade is only as long as the switch takes

 Disadvantages:

• implementation
• it needs constant supervision, I haven't seen any single SW replication  which run completely autonomous
• Journal implementation must be well designed, you have to review all sw configuration if still all objects are really replicated
• Files synchronization may take ages.  

HW replication:

Advantages: 

• Almost no implementation costs, no application changes
• Very east for maintenance and management (set and forget)
• No additional OS licenses needed

 Disadvantages:

• Each write I/O needs to travel to remote location ( make sure you have good connection)
• Replicate Temporary storage
• DR machine permanently powered off, can't be used for tests, reporting
• OS upgrade requires regular outage
• Requires external storage and SAN
• Abnormal IPL when system crash. - you replicate the entire OS, if the production is lost and recovery is happening on DR node, The OS must go thru entire Database recovery

From my personal experience, of course it depends what is your priorties. I can say that everyone goes to HW replication mainly due to issues and time needed to solve SW replication problems. If you have very important application running and you want to make sure that almost all the time you are ready to switch. You need a full time employee to monitor and react on replication issues. This overhead you don't have with HW replication

------------------------------
Bartlomiej Grabowski
IBM Champion - Platinum Redbook Author and Principal System Specialist
------------------------------

Hi Michael

i'm using FSR solution in our shared infrastructure for many customers. In my opinion it is simple (installation, configuration and administration) ,  and reliable tool, which can help you in switching to backup location in case of any planned maintenance services or hardware failure on your primary site. Many times we use it before patching production systems (there is no replication during that time), so it allows to "restore" your system from backup site in case of any issue.
Obviously, it is based on external storage replication so everything is replicated to target system. 
Regarding serial numbers for applications, usually customers are prepared and already have license keys also for target system, but is probably normal situation when you have DR solution between two servers.

Hi Michael - there are some really good points made here. I might just add a couple more…

A Power 10 in combination with FS7300 gives you fantastic resilience and availability options.

To add a couple of further advantages for MIMIX from Bartlomiej:

MIMIX provides a quicker RTO than full system replication as the target OS already available. Depending on the time of failure, an unplanned DR invocation of FSR may lead to an extended abnormal IPL time. Depending on network bandwidth, FSR may have a longer RPO than MIMIX as well.

If you have a future plan to migrate to IBM Power VS then FSR will not help you get there, MIMIX will & with minimal impact.

FSR can enable hardware migrations, but you need additional storage or to forego DR while it's reconfigured.

MIMIX can automate a very quick switch and therefore it provides an easy way of applying OS upgrades, PTFs and future hardware migrations.

MIMIX can be paused at a specific point easier than FSR. You can automatically pause MIMIX exactly before a batch run and restart afterwards, providing an easy  way to set an application aware RPO.

With reference to the FS7300, it provides enhanced operational resilience functionality which is compatible with both hardware and software replication. Traditional Flashcopy has been superseded by policy based snapshots - immutable backups called Safeguarded Copy. This can be used to create a modern recovery strategy that not only covers 'traditional' DR but also recovery from cyber attack or data corruption.

We are also using Safeguarded Copy.

Hi Michael - there are some really good points made here. I might just add a couple more…

A Power 10 in combination with FS7300 gives you fantastic resilience and availability options.

To add a couple of further advantages for MIMIX from Bartlomiej:

MIMIX provides a quicker RTO than full system replication as the target OS already available. Depending on the time of failure, an unplanned DR invocation of FSR may lead to an extended abnormal IPL time. Depending on network bandwidth, FSR may have a longer RPO than MIMIX as well.

If you have a future plan to migrate to IBM Power VS then FSR will not help you get there, MIMIX will & with minimal impact.

FSR can enable hardware migrations, but you need additional storage or to forego DR while it's reconfigured.

MIMIX can automate a very quick switch and therefore it provides an easy way of applying OS upgrades, PTFs and future hardware migrations.

MIMIX can be paused at a specific point easier than FSR. You can automatically pause MIMIX exactly before a batch run and restart afterwards, providing an easy  way to set an application aware RPO.

With reference to the FS7300, it provides enhanced operational resilience functionality which is compatible with both hardware and software replication. Traditional Flashcopy has been superseded by policy based snapshots - immutable backups called Safeguarded Copy. This can be used to create a modern recovery strategy that not only covers 'traditional' DR but also recovery from cyber attack or data corruption.

Safeguarded copy of on my checklist of things I'd like to implement on our FS7300s. 

We are also using Safeguarded Copy.

Hi Michael - there are some really good points made here. I might just add a couple more…

A Power 10 in combination with FS7300 gives you fantastic resilience and availability options.

To add a couple of further advantages for MIMIX from Bartlomiej:

MIMIX provides a quicker RTO than full system replication as the target OS already available. Depending on the time of failure, an unplanned DR invocation of FSR may lead to an extended abnormal IPL time. Depending on network bandwidth, FSR may have a longer RPO than MIMIX as well.

If you have a future plan to migrate to IBM Power VS then FSR will not help you get there, MIMIX will & with minimal impact.

FSR can enable hardware migrations, but you need additional storage or to forego DR while it's reconfigured.

MIMIX can automate a very quick switch and therefore it provides an easy way of applying OS upgrades, PTFs and future hardware migrations.

MIMIX can be paused at a specific point easier than FSR. You can automatically pause MIMIX exactly before a batch run and restart afterwards, providing an easy  way to set an application aware RPO.

With reference to the FS7300, it provides enhanced operational resilience functionality which is compatible with both hardware and software replication. Traditional Flashcopy has been superseded by policy based snapshots - immutable backups called Safeguarded Copy. This can be used to create a modern recovery strategy that not only covers 'traditional' DR but also recovery from cyber attack or data corruption.

We have the SafeGuarded Copy Toolkit for IBM i.  Allows us to schedule copies from IBM i.  IBM Lab helped us set it up.

Safeguarded copy of on my checklist of things I'd like to implement on our FS7300s. 

We are also using Safeguarded Copy.

Hi Michael - there are some really good points made here. I might just add a couple more…

A Power 10 in combination with FS7300 gives you fantastic resilience and availability options.

To add a couple of further advantages for MIMIX from Bartlomiej:

MIMIX provides a quicker RTO than full system replication as the target OS already available. Depending on the time of failure, an unplanned DR invocation of FSR may lead to an extended abnormal IPL time. Depending on network bandwidth, FSR may have a longer RPO than MIMIX as well.

If you have a future plan to migrate to IBM Power VS then FSR will not help you get there, MIMIX will & with minimal impact.

FSR can enable hardware migrations, but you need additional storage or to forego DR while it's reconfigured.

MIMIX can automate a very quick switch and therefore it provides an easy way of applying OS upgrades, PTFs and future hardware migrations.

MIMIX can be paused at a specific point easier than FSR. You can automatically pause MIMIX exactly before a batch run and restart afterwards, providing an easy  way to set an application aware RPO.

With reference to the FS7300, it provides enhanced operational resilience functionality which is compatible with both hardware and software replication. Traditional Flashcopy has been superseded by policy based snapshots - immutable backups called Safeguarded Copy. This can be used to create a modern recovery strategy that not only covers 'traditional' DR but also recovery from cyber attack or data corruption.

That's my plan is to have IBM Labs help us set that up.  

------------------------------
Michael Soucy
Sr Midrange Engineer
Bangor Savings Bank
Bangor ME
------------------------------


Original Message:
Sent: Mon March 30, 2026 02:42 PM
From: Robert Berendt
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

We have the SafeGuarded Copy Toolkit for IBM i.  Allows us to schedule copies from IBM i.  IBM Lab helped us set it up.

------------------------------
Robert Berendt IBMChampion
Business Systems Analyst, Lead
Dekko
Fort Wayne


Original Message:
Sent: Mon March 30, 2026 02:31 PM
From: Michael Soucy
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

Safeguarded copy of on my checklist of things I'd like to implement on our FS7300s. 

------------------------------
Michael Soucy
Sr Midrange Engineer
Bangor Savings Bank
Bangor ME


Original Message:
Sent: Mon March 30, 2026 07:36 AM
From: Robert Berendt
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

We are also using Safeguarded Copy.

------------------------------
Robert Berendt IBMChampion
Business Systems Analyst, Lead
Dekko
Fort Wayne


Original Message:
Sent: Sun March 29, 2026 06:14 PM
From: Edward Yeates
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

Hi Michael - there are some really good points made here. I might just add a couple more…

A Power 10 in combination with FS7300 gives you fantastic resilience and availability options.

To add a couple of further advantages for MIMIX from Bartlomiej:

MIMIX provides a quicker RTO than full system replication as the target OS already available. Depending on the time of failure, an unplanned DR invocation of FSR may lead to an extended abnormal IPL time. Depending on network bandwidth, FSR may have a longer RPO than MIMIX as well.

If you have a future plan to migrate to IBM Power VS then FSR will not help you get there, MIMIX will & with minimal impact.

FSR can enable hardware migrations, but you need additional storage or to forego DR while it's reconfigured.

MIMIX can automate a very quick switch and therefore it provides an easy way of applying OS upgrades, PTFs and future hardware migrations.

MIMIX can be paused at a specific point easier than FSR. You can automatically pause MIMIX exactly before a batch run and restart afterwards, providing an easy  way to set an application aware RPO.

With reference to the FS7300, it provides enhanced operational resilience functionality which is compatible with both hardware and software replication. Traditional Flashcopy has been superseded by policy based snapshots - immutable backups called Safeguarded Copy. This can be used to create a modern recovery strategy that not only covers 'traditional' DR but also recovery from cyber attack or data corruption.

------------------------------
Edward Yeates
Professional Services Manager

Hi Michael - if you are at Common Power Up 26, you can attend my technical session 'The IBM Power Clean Room' (April 30th) where I will explain how you can set-up a clean room partition to help validate and recover safeguarded copy snapshots as part of enhanced IBM i operational resilience.

PU26 - Education - Common 

------------------------------
Edward Yeates
Professional Services Manager

Celerity Ltd
------------------------------


Original Message:
Sent: Mon March 30, 2026 03:05 PM
From: Michael Soucy
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

That's my plan is to have IBM Labs help us set that up.  

------------------------------
Michael Soucy
Sr Midrange Engineer
Bangor Savings Bank
Bangor ME


Original Message:
Sent: Mon March 30, 2026 02:42 PM
From: Robert Berendt
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

We have the SafeGuarded Copy Toolkit for IBM i.  Allows us to schedule copies from IBM i.  IBM Lab helped us set it up.

------------------------------
Robert Berendt IBMChampion
Business Systems Analyst, Lead
Dekko
Fort Wayne


Original Message:
Sent: Mon March 30, 2026 02:31 PM
From: Michael Soucy
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

Safeguarded copy of on my checklist of things I'd like to implement on our FS7300s. 

------------------------------
Michael Soucy
Sr Midrange Engineer
Bangor Savings Bank
Bangor ME


Original Message:
Sent: Mon March 30, 2026 07:36 AM
From: Robert Berendt
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

We are also using Safeguarded Copy.

------------------------------
Robert Berendt IBMChampion
Business Systems Analyst, Lead
Dekko
Fort Wayne


Original Message:
Sent: Sun March 29, 2026 06:14 PM
From: Edward Yeates
Subject: Questions about PowerHA Tools for IBM i - Full System Replication

Hi Michael - there are some really good points made here. I might just add a couple more…

A Power 10 in combination with FS7300 gives you fantastic resilience and availability options.

To add a couple of further advantages for MIMIX from Bartlomiej:

MIMIX provides a quicker RTO than full system replication as the target OS already available. Depending on the time of failure, an unplanned DR invocation of FSR may lead to an extended abnormal IPL time. Depending on network bandwidth, FSR may have a longer RPO than MIMIX as well.

If you have a future plan to migrate to IBM Power VS then FSR will not help you get there, MIMIX will & with minimal impact.

FSR can enable hardware migrations, but you need additional storage or to forego DR while it's reconfigured.

MIMIX can automate a very quick switch and therefore it provides an easy way of applying OS upgrades, PTFs and future hardware migrations.

MIMIX can be paused at a specific point easier than FSR. You can automatically pause MIMIX exactly before a batch run and restart afterwards, providing an easy  way to set an application aware RPO.

With reference to the FS7300, it provides enhanced operational resilience functionality which is compatible with both hardware and software replication. Traditional Flashcopy has been superseded by policy based snapshots - immutable backups called Safeguarded Copy. This can be used to create a modern recovery strategy that not only covers 'traditional' DR but also recovery from cyber attack or data corruption.

------------------------------
Edward Yeates
Professional Services Manager

This is all valuable information.  I have a lot of things to consider.  I don't know if this makes any difference, but the LPAR we want to replicate is our testing / development environment for Jack Henry.  The Jack Henry data is stale as of a certain point in time.  We do run a Jack Henry end of day on that partition, so files are getting updated. However, we do refresh the data at least two or three times a year using a backup of the data library from production. That process involves deleting records, and then copying data from one library to another. There is minimal to moderate usage of this partition, but nowhere the amount of traffic as our production environment. 

I do like MIMIX for our development environment. The ability to pause replication is utilized when we do our EOY processing. At one point in that process, we have to stop replication so our Data & Analytics team can run their Cognos reports. Cognos pulls data from the target partition. It protects production from having to deal with runaway Cognos reports that eat up CPU. 

Precisely reached out to me to schedule a meeting with us on Wednesday to go over the pros and cons of using either software replication or hardware replication. I'm anxious to hear what they have to say.

Thank you again for all of this information.

-Mike

This is all valuable information.  I have a lot of things to consider.  I don't know if this makes any difference, but the LPAR we want to replicate is our testing / development environment for Jack Henry.  The Jack Henry data is stale as of a certain point in time.  We do run a Jack Henry end of day on that partition, so files are getting updated. However, we do refresh the data at least two or three times a year using a backup of the data library from production. That process involves deleting records, and then copying data from one library to another. There is minimal to moderate usage of this partition, but nowhere the amount of traffic as our production environment. 

I do like MIMIX for our development environment. The ability to pause replication is utilized when we do our EOY processing. At one point in that process, we have to stop replication so our Data & Analytics team can run their Cognos reports. Cognos pulls data from the target partition. It protects production from having to deal with runaway Cognos reports that eat up CPU. 

Precisely reached out to me to schedule a meeting with us on Wednesday to go over the pros and cons of using either software replication or hardware replication. I'm anxious to hear what they have to say.

Thank you again for all of this information.

-Mike

Typo in my last reply.  I said, "I do like MIMIX for our development environment.", but what I meant to say was "I do like MIMIX for our production environment".  I just wanted to clarify that. I think we have found two banks that are using the FSR tool so it's going to be interesting to compare their experience to what we do know with MIMIX on our production environment.  Mimix was setup in our production environment with the help from Jack Henry many years ago before I even started working at BSB. 

This is all valuable information.  I have a lot of things to consider.  I don't know if this makes any difference, but the LPAR we want to replicate is our testing / development environment for Jack Henry.  The Jack Henry data is stale as of a certain point in time.  We do run a Jack Henry end of day on that partition, so files are getting updated. However, we do refresh the data at least two or three times a year using a backup of the data library from production. That process involves deleting records, and then copying data from one library to another. There is minimal to moderate usage of this partition, but nowhere the amount of traffic as our production environment. 

I do like MIMIX for our development environment. The ability to pause replication is utilized when we do our EOY processing. At one point in that process, we have to stop replication so our Data & Analytics team can run their Cognos reports. Cognos pulls data from the target partition. It protects production from having to deal with runaway Cognos reports that eat up CPU. 

Precisely reached out to me to schedule a meeting with us on Wednesday to go over the pros and cons of using either software replication or hardware replication. I'm anxious to hear what they have to say.

Thank you again for all of this information.

-Mike

Both have their pro's and con's but using MiMiX in a modern environment with various webservers, SSL configurations, application definitions, ...  can be a challenge to replicate.

Rule number one with MiMiX is therefore to test your switch framework at a regular base while with storage replication it is set and forget.

Typo in my last reply.  I said, "I do like MIMIX for our development environment.", but what I meant to say was "I do like MIMIX for our production environment".  I just wanted to clarify that. I think we have found two banks that are using the FSR tool so it's going to be interesting to compare their experience to what we do know with MIMIX on our production environment.  Mimix was setup in our production environment with the help from Jack Henry many years ago before I even started working at BSB. 

This is all valuable information.  I have a lot of things to consider.  I don't know if this makes any difference, but the LPAR we want to replicate is our testing / development environment for Jack Henry.  The Jack Henry data is stale as of a certain point in time.  We do run a Jack Henry end of day on that partition, so files are getting updated. However, we do refresh the data at least two or three times a year using a backup of the data library from production. That process involves deleting records, and then copying data from one library to another. There is minimal to moderate usage of this partition, but nowhere the amount of traffic as our production environment. 

I do like MIMIX for our development environment. The ability to pause replication is utilized when we do our EOY processing. At one point in that process, we have to stop replication so our Data & Analytics team can run their Cognos reports. Cognos pulls data from the target partition. It protects production from having to deal with runaway Cognos reports that eat up CPU. 

Precisely reached out to me to schedule a meeting with us on Wednesday to go over the pros and cons of using either software replication or hardware replication. I'm anxious to hear what they have to say.

Thank you again for all of this information.

-Mike

There has been a lot of good information in this thread about both FSR and logical replication products.  Two statements I'd really like to draw your attention to are:

-Rich Malloy

-Paul Nicolay

I don't quite agree with the "set and forget" part.  FSR certainly does not require constant maintenance and monitoring in the same way logical replication products do, but it still should be tested and monitored on a regular basis since you business relies on it.

I'm going to say "consider the source."  The vendor (myself included since I'm an IBM Expert Labs HA/DR consultant) is going to tell you things that make their solution sound better.

For any DR solution, consider RPO and RTO.  RPO is basically how much data you can afford to lose for recovery, RTO is how long you expect it to take to recover from a downtime.  Don't completely trust the vendors (even me) - ask the reference customers about their actual experiences.  If you are interested in DR, specifically ask:

If they have done a planned role swap - how long did it take?

If not, why not? 

If they have tested an unplanned role swap - starting from a DR instance without any preparation - How long did it take (RTO)?  How much data did they lose (RPO)?    I should note that talking to a safeguarded copy customer that has mounted a copy on a recovery partition would have executed this test regularly - ask them how long it takes to start the recovery partition.

Do they use the product to do regular A/B roles swaps between sites? -- this is the gold standard of a DR solution - the best test is always to run production in DR, and any backup that has not been tested is NOT a reliable backup.

There are plenty of customer that use both storage and logical replication, for different purposes.  You may be one of those customers. 

Both have their pro's and con's but using MiMiX in a modern environment with various webservers, SSL configurations, application definitions, ...  can be a challenge to replicate.

Rule number one with MiMiX is therefore to test your switch framework at a regular base while with storage replication it is set and forget.

Typo in my last reply.  I said, "I do like MIMIX for our development environment.", but what I meant to say was "I do like MIMIX for our production environment".  I just wanted to clarify that. I think we have found two banks that are using the FSR tool so it's going to be interesting to compare their experience to what we do know with MIMIX on our production environment.  Mimix was setup in our production environment with the help from Jack Henry many years ago before I even started working at BSB. 

This is all valuable information.  I have a lot of things to consider.  I don't know if this makes any difference, but the LPAR we want to replicate is our testing / development environment for Jack Henry.  The Jack Henry data is stale as of a certain point in time.  We do run a Jack Henry end of day on that partition, so files are getting updated. However, we do refresh the data at least two or three times a year using a backup of the data library from production. That process involves deleting records, and then copying data from one library to another. There is minimal to moderate usage of this partition, but nowhere the amount of traffic as our production environment. 

I do like MIMIX for our development environment. The ability to pause replication is utilized when we do our EOY processing. At one point in that process, we have to stop replication so our Data & Analytics team can run their Cognos reports. Cognos pulls data from the target partition. It protects production from having to deal with runaway Cognos reports that eat up CPU. 

Precisely reached out to me to schedule a meeting with us on Wednesday to go over the pros and cons of using either software replication or hardware replication. I'm anxious to hear what they have to say.

Thank you again for all of this information.

-Mike