We are Apps 9 UPD13 and opted to go to Cloud skipping V10 because we didn’t want to do an upgrade followed by another upgrade a year or so later to stay within the maintenance restrictions that have been imposed.  We decided to do it all at once because the interim upgrade to V10 gains you nothing but time, maybe the chance to learn some of Aurena.  However, there are limited Aurena configuration tools in V10, the full suite only appears in Cloud so from the perspective of getting your processes and customizations converted to the Aurena version, we found that path to be of almost no value for our situation.  That is a very superficial explanation of why we skipped, it was actually months of evaluation when there was nothing available to guide the decision.

From the Cloud perspective, you have to make some very harsh assumptions to prepare yourself for the difficulty that is about to come, doesn’t matter if you are coming from V9 or V10.  These were my assumptions and to do date they have not been wrong, in fact, they weren’t harsh enough in some areas.

1. Cloud was not designed with upgrade customers in mind.  It is so different in look, feel and operation, no matter how experienced you are with IEE, it is a struggle to do anything.
2. It is slow.  Not the system or application, but the number of steps to do the same process in Cloud is in every case more clicks to do the same thing.  If you are a keyboard operator and not a mouse person, it is even worse.
3. You should assume that every customization you have outside of events will not uplift and will have to be touched if not completely rewritten to perform the same function.  This is a huge task for us as well.
4. If you are using anything but bare bones IFS, you are going to have problems.  Many components and modules are still being released (Global Extension, certain localizations, and other things) or have limited functionality (Query Builder Tool).
5. Some existing interfaces (Vertex Tax, Credit Card, Freight Interface) are so different there isn’t an upgrade path from point A to B, it is a new deployment and migration for those components with additional cost and third party software required.
6. IFS haven’t done or haven’t figured many of these things yet themselves on an upgrade, so prepare to be an experiment.
7. Many of the tools you may have depended on to build your process in V9 IEE simply don’t exist in Cloud (SQL Query Tool, Right Mouse Button ability to execute code, many others that are listed in other threads about the differences between IEE and Aurena), so we don’t even know what the like for like replacement is for our customizations yet.
8. The idea of getting a problem patched no longer exists, everything is a 5 week delivery minimum for the next service pack, that is only useful if your bug fix is in that next delivery.  Cloud is very buggy and has been through the 21 series releases, 22R1 is supposed to be better, but it was just released so we haven’t installed it yet.  We knew going in that we can’t go live before at least 22R2 is released, but even that is not a guarantee for us yet.
9. Detailed documentation for everything is lagging behind the release of the modules and functions, there is superficial information and presentations of what is new, but in truth very little of that has been useful for upgrade customers looking to get back to where they were.
10. The entire upgrade process front end vetting has been extremely frustrating to put it mildly, it is definitely something IFS is figuring out as they go and not something that they baked into the plan for the Cloud release.
11. The Build Place, delivery process, VPN tunnel, hosting arrangement is significantly complex, not well explained, and most likely significantly more expensive from a server space and performance perspective than you might imagine.  V9 IEE requirements for servers and Cloud Aurena requirements are night and day.
12. If you are going to do any work in house, you need a very strong programmer/Db Admin or else you are at the mercy of IFS doing the work and it won’t be cost effective.
13. Even after all of that, you should also expect that all of your permissions have to be re-done, there is no upgrade path for that.  All of your screen adjustments for efficiency or clarity have to be re-done, in fact they likely have to be re-done on each subsequent update once you start taking them.  Evergreen is not at all the smooth solution it is advertised to be.
14. I could go on, and on….

To my knowledge….no one has completed an upgrade from V9 to Cloud yet, meaning live and operational, we are one of a few in progress as I understand it.

Do I regret jumping in early to the upgrade process to Cloud, not yet, but that isn’t to say there haven’t been lots of reasons to regret it.

Is the process smooth and transparent?  Not in the least.

Do I have any idea when we will go live?  There is a target of a year from now, but so far we’ve only uncovered loads more work to do, in no case has there been less work than I anticipated, so I am maybe only 50% certain one year is achievable.

Do your homework.  Ask lots of questions, demand lots of answers.  Build a very strong team if you can to do as much work internally, or be prepared to spend a lot more than you expect if you have to have the work from the outside.

Eyes wide open and double and triple check everything you are told because the likelihood is that it isn’t the full version of the facts.

We are on v9 UPD16. We decided to go via v10 for a few reasons. We wanted the quickest easiest path back to a supported system. We would like to prepare some of our interfaces with the use of the new RestApi’s. We wanted to keep EE, and drip feed departments over to Aurena in a controlled manner. We felt Cloud wasn’t quite ready and a few unknowns still existed for us, despite the lure of not having to do it all again a couple of years later….

Sounds like Cloud is ok if you have an out of the box setup, which I dont think we are far from.

Cheers

Mike

Thanks for the useful inputs @ShawnBerk and @MikeArbon. I will take these up for our internal discussions. 

We’ve been told Apps 10 upgrade from Apps 9 is not being offered unless you have already started the journey. We are on Apps 9 - Update 17. 

Reading @ShawnBerk’s comments I’m going to have some sleepless nights. Hopefully things will be smoother by the time we start the journey. Hoping to complete in 9 months from Placement of PO. 

Hoping to complete in 9 months from Placement of PO. 

For context, it will be 6 months from the point of PO placement for us until we have a fully running system that we can then BEGIN the process of recreating permissions, uplifting customizations, and creating new contexts.  There was also 6 months of pre-work before we placed the PO. In total, it will be 22 months if we make the current schedule, from when we started.

Don’t underestimate the front end.

Hi 
App9 to IFS cloud upgrade is technicly Possible. However IF you have lots of customization with Aurena clinent that will take some time. Some customer are planning to go to 2 stage (APP9 to APP10 and IFS cloud)and i worked with them too. That is possible but it will take more time. 

@Maharawaththa 

What about this comment from above?  Can you confirm this is now policy for customers?

We’ve been told Apps 10 upgrade from Apps 9 is not being offered unless you have already started the journey. We are on Apps 9 - Update 17. 

@ShawnBerk 

Since App10 RND support will end soon, it is  therotically true. Howerver as technical service we are handling the App10 Upgrade and estimation as well,  if customer request. Customer has to know the limitation of reach the RND. 

The key there is the word "client". If your old customizations include changes to the IEE forms or logic in the clients, then some amount of work must be done. Some kind of client customizations might not even be possible to do in Aurena since the two platforms are so different on the client side.  Then you need to rethink the whole customization, to fit how Aurena is built. On the positive side, customizations that fits well with Aurena are probably cheaper to develop and much much faster to test (the development cycle is really fast and a programmer and "functional person" can get things done really fast).

For customizations to business logic, uplifting them to Aurena should not be harder than to uplift them from Apps 9 to Apps 10. It's similar to uplifting custom events, which should generally work without any issues, or at least no more issues than between any Apps upgrade.
 

Yes, things like….

• Custom APIs called from Custom RMBs
• Quick Reports built from a SQL Query
• Performing adhoc queries using the SQL Query tool at the table level
• Pretty much anything requiring the Global Extension
• IFSNA Freight Interface
• Vertex Tax Interface
• China Localization
• Reports or searches that return anything more than 24 records
• Tasks dependent on migration jobs
• Sub-query SQL searches

Essentially anything that took advantage of the many tools of IEE to make things quicker and efficient either can’t be replicated or have to be re-created in a whole new way if they can be.  Anything beyond a dead simple installation using only core level functions is going to be difficult or not possible.  Still looking for the things that are easier and quicker, they are quite elusive.

Thanks all for valuable input, unfortunately I cannot select a best answer since I started the topic in ‘conversation’ mode, not ‘question’ mode. 

One question for @ShawnBerk, did your company manage to get a cloud demo environment from IFS to study the new features, etc? 

No worries on the best answer, this is a topic that there is no one answer for everyone, it is a hard decision that is company by company.

@Asela Munasinghe 

We delayed our decision for over 6 months with the lack of a RACE or demo environment being the main factor.  We had several live demos with our sales engineer where he would drive and our whole team would ask questions about this and that to explore.  It was their sales environment.  Then finally, we were able to get 3 of our team separate access to their internal environment, but it was significantly painful from an access standpoint and still had restrictions.  We could explore, but couldn’t customize or change too many things.  It was enough for it to be the tipping point for us to believe the software was capable of running our business eventually (not withstanding a mountain of work), but the devil really is in the details as we’re now finding out.

IFS is rethinking the whole decision to NOT have a RACE environment, but the Azure Web Service hosting and security around opening up a cloud application to the world has some very significant challenges.

It took lots of arm twisting and ranting to get to the point we did with the demo environment, so all I can say is continue to voice how important that is to the decision making process.

We were advised to go from Apps 9 to Apps 10 because it wouldn’t be a culture shock in moving to the Aurena front end immediately… and from how the upgrade is going we are glad we made that decision.
 

@MikeArbon when you say ‘out of the box setup’ what exactly do you mean? Is it similar to a setup that has quite a lot of customizations and well as integrations with other systems? 

@darylslinn , was it your organization’s or IFS’s recommendation to go through App10? 

It was suggested as an option for time and potential cost saving for the first part of the overall upgrade project.

We are currently planning on starting a Apps9upd17->Apps10upd15 project. The 10-upgrade was first thought around 2019 but it did not provide anything special then so we have done two Apps9 sub-version updates since then instead (you could argue were they worth it). Main driver for the 10-upgarade is keeping up in supported versions and trying to minimize the tech debt. During the Apps10 usage the aim is to do step-by-step integration renewals and transitions to the new UI.

The remaining supported life of 6 years until 2028 for Apps10 is not that long when you wonder the business case for the total implementation projects and related work and trouble. Reading the above thread about all the challenges the Cloud brings along it really makes one think is this work with a meaning .

Hi Guys,

We are in the middle of the storm (IFS 9 to Cloud 22R1). It’s been an emotional roller coaster. IFS Cloud is far from been ready. They sold like everything was ready to go (and we have not a lot of customizations), so we decide to go direct to IFS CLOUD. Below, some points:

1. IFS Cloud is not ready
2. So many bugs
3. Worst support ever
   1. Any one of our issues had solutions from IFS. My team and I found the solutions (some with your help guys).

4. We have a case open for 3 months now, and any solution has been given.
5. Focus in advanced functions and let many bugs on basic functions.

This list can go far, but the point is, wherever your decision is, your team will be the key of success. Don’t wait for IFS, your expectations need to be as low as possible, frustration, angry and sometimes a little bit of joy.

PS. I don’t know if some of you will be at IFS UNLISHED, but if you want to talk, it will be very interesting share our experiences. 

Gianni Neves 

Just re-read this thread. We have recently started the upgrade from 9 to 10 project that I mentioned above and the golive date in prod is almost a year from now. We have a bunch of customizations (n ~200).

Our biggest reasons not to jump to IFS Cloud directly:

A) upgrading the customizations without having to re-plan every process at once (which would have required waay too much effort. It would be a partial re-implementation with changes needed to many e2e processes and also to other connecting systems).

B) continue using IEE to avoid a big bang project

C) agreement related reasons

We have tried to “cut the elephant to smaller pieces” by doing a technical code upgrade first and acknowledge that there are many continuation projects to do when first in IFS10. These include the integration technique change to using rest api’s, the switch to Aurena UI and some internal component switchovers from older to newer component (like budgeting functionalities).

At the same time it already worries a lot thinking about the possible later Cloud project. How can you have a reasonable business case to justify the project financially? How can you get rid of your customizations if the result is that effectiveness suffers and manual work increases. It sounds like reinventing the wheel when you have to re-implement all your processes (A on the list above) moving from 10 to Cloud.

Hi @ShawnBerk and @Asela Munasinghe ,

I’ve been following the IFS Application 10 to IFS Cloud upgrade discussions carefully because we have been thinking of upgrading to IFS Cloud potentially in 4Q 2023.   We keep all our servers on-premise and we are looking at replacing our servers in 2023 and I’d like to account for IFS Cloud as part of the server analysis we will be doing in the coming weeks and months.   

Do either of you have real world IFS Cloud performance and server configurations as part of your testing?   I’m concerned about memory, processor and network band width on the application server and database servers in particular.   IFS Cloud changing the application server equation with Aurena and I’m assuming the application server will require more resources to provide similar performance to using IEE.

Any information you have would be of help.

Regards,

William Klotz

@william.klotz 

I’ll see what I can find that I can share.

Expect that whatever you had for IEE, needs to be twice as good and you can no longer share servers as we did with our non-Prod environments previously.  It is three servers per environment and if they arent close to the recommended specs, performance with Cloud is abysmal.

The only server we are sharing right now it is the DB non-prod. Like @ShawnBerk told, all other servers need to be separated.

Our go live is planned for begin December, so we are working already on memory, working and processor optimization.

Thanks,

Gianni Neves

Interesting read...

We are currently looking to upgrade from IFS apps 9 to on prem IFS cloud within the next 4 months... 

We have a prod and dev environment 2 servers each (app server and database server)

How flexible are the licensing with cloud, is it like MS Office licenses - scale up and down as you need?

Hope that’s start of project, not time-line!

We were sold licenses based on a 5 year term - also bare in mind what you currently have on perpetual, you may need to pay a uplift in support for them. You’ve got to be careful and is not a 1:1 mapping between Apps9 and Cloud licensing modules - accidents can happen and you could nearly be charged for what you already ‘own.’

It is a minefield and caused more than one ‘death by spreadsheet’ moment.

Hi @ShawnBerk. We are embarking on a similar journey (Apps 9 UPD17 to Cloud). Thank you so much for providing the original 14 points from 9 months ago. I’m wondering if you would be willing to provide an update on how the project has gone for you. Some of the things that you have learned, and especially anything that we should be looking at from an API / customization perspective. Again - thank you so much for your contributions to date!