Question

Upgradeable Transformers and FME Server

  • 16 October 2019
  • 4 replies
  • 19 views

Badge +10

Hello,

Can someone please explain the "behaviour" of an upgradeable transformer (Desktop 2019) in relation to FME Server 2019 when the workspace gets submitted?

So let's say you have 2 TestFilters, same version.

You upgrade one but the other still remains on the current version within the Workspace.

 

How is that possible?

 

Are each transformers considered as "isolated entities"?

What happens then when the workspace is published to FME Server?

 

Are the transformers automatically upgraded or do I end up with different versions of the same?

 

(I won't be able to upgrade anything at that point unless I download it and re-publish it again)

That leads me to my final question:

If a transformer version becomes deprecated how would FME Server behave then?

 

Will it carry on and ignore or return an error?

 

Many thanks

 

Sameer

4 replies

Userlevel 5
Badge +25

Yes, the transformer versions exist as separate entities with the FME engine, so on desktop and server and obviously old FME versions don't know about new ones. So as long as your workspace was developed with a build number that's lower or the same as the one you're running on FME Server there should be no problem, even if they contain transformers that are deprecated (or even multiple versions of the same transformer in one workspace). FME will not automatically upgrade transformers for you (as you've seen on desktop there may be a need to look at the parameters again and only a human operator is able to set those properly).

Having said that, I would recommend upgrading all transformers before uploading to FME Server, just to be on the safe side. There's a few circumstances where, in theory, weird behaviour could occur if you have different transformer versions.

 

Userlevel 4
Badge +25

A great question, and great answer from @redgeographics

I just wanted to add that the key for Safe is backwards compatibility: your workspace should run the same as it always has done, regardless of platform. We'll never upgrade a transformer automatically. And if a transformer is deprecated, you can't add a new instance, but the old one should keep on going.

There are rare exceptions, for example Python 2.7 will stop working soon, because the underlying product is being deprecated. The same with Google Fusion Tables.

Badge +10

@redgeographics, @mark2atsafe

Many thanks for the prompt reply with clear and concise answers. I shall bear this in mind in the future. In fact that might be a good opportunity to "revisit" the Workspaces albeit time consuming.

My only suggestion would be:

To be able to upgrade the same transformers within an existing Workspace at once. Not sure if that option would be feasible.

All the best

 

Sameer
Userlevel 4
Badge +25

@redgeographics, @mark2atsafe

Many thanks for the prompt reply with clear and concise answers. I shall bear this in mind in the future. In fact that might be a good opportunity to "revisit" the Workspaces albeit time consuming.

My only suggestion would be:

To be able to upgrade the same transformers within an existing Workspace at once. Not sure if that option would be feasible.

All the best

 

Sameer

Hi Sameer,

Yes I think the ability to upgrade all the same transformers at once is on our list of things to do.

Reply