Sure, we will take a note of it to improve the diff usage help for more clarity.
Original Message:
Sent: Aug 29, 2024 06:02 AM
From: Michael Mueller
Subject: graphman-client diff
@Raju Gurram
Hi Raju.
As a comment:
It might be just myself, but, it wasn't obvious/understandable, that the diff "include..." options are intended exclusively for the "input-report" function, as far as I can see.
And that two of them (includeUpdates, includeInserts) are defaulted to true while the includeDeletes is defaulted to false.
Maybe a better & complete description through the "diff --help" might improve the overall understanding.
Thanks
...Michael
Original Message:
Sent: Aug 29, 2024 02:20 AM
From: Michael Mueller
Subject: graphman-client diff
@Raju Gurram
Thanks a million, I wasn't aware of the --input-report capability and its purpose. That helps and answers all my questions (for now :)
Kind regards
...Michael
Original Message:
Sent: Aug 29, 2024 12:27 AM
From: Raju Gurram
Subject: graphman-client diff
@Michael Mueller
Having the diff-report in-place, you could ask the diff command to re-generate the diff-bundle anytime.
./graphman.sh --input-report <diff-report> --output <diff-bundle>
For more information, please refer to the diff command help (./graphman.sh diff --help).
As you specified, if anything is not meeting the requirements, it is encourage to make use of the client's extension points.
Thanks,
Raju
Original Message:
Sent: Aug 28, 2024 07:57 AM
From: Michael Mueller
Subject: graphman-client diff
@Raju Gurram
Thanks Raju for clarification.
Makes sense !
However: In the past, we've updated the mapping data between the "diff" and the "revise" command execution for a special customer purpose. (This was and is necessary to map generic_entity goid references between different gateways.
The "only" supported way doing the same, without changing ootb js code, now, is through the pre-import extension, as far as I am aware.
I know, that generic entities are now supported by gw v11.1.1. Customer is still on v11.0.02. So I am wondering if there is any other capability to influence these mappings ?
Any idea?
Thanks.
...Michael
Original Message:
Sent: Aug 28, 2024 06:38 AM
From: Raju Gurram
Subject: graphman-client diff
@Michael Mueller
Your understanding is 100% correct w.r.t. diff and revise commands. As part of their evolution, responsibilities are exchanged for good reasons.
Earlier, revise command was used to complement the bundle during the import operation by matching goid/guid in the target environment. The diff command is essential in identifying the differences between source and target environments. As part of which, it collects and embeds the machine identifier (goid, guid) differences inside the bundle.
With the v1.3 release, the diff command is enhanced to generate the detailed difference report. Now it becomes a place-holder machine identifier (goid/guid) differences as well. I believe, this revised decision not only separates the standard bundle from the intermediate details, but also improves the user-experience while working with the commands. The revise command is now dedicated to migrate the bundles from one schema to another.
Thanks,
Raju
Original Message:
Sent: Aug 27, 2024 02:49 AM
From: Michael Mueller
Subject: graphman-client diff
Dear Team.
I am looking for confirmation of my understanding of the current graphman-client version 1.3.00
My question is about how the client handles differences of goid's/guid's for different gateways for the same entity/resource.
Previously, before 1.3, this data was collected through the diff command and listed in goidMapping/guidMapping structures in a bundle.
We have used the revise command to do the actual replacements of source goid/guids by the target ones.
With version 1.3 the actual replacements are done by the diff command directly.
Correct? Always ?
The revise command "only" takes care of migrating "old" entityNames to new ones , e.g. fipUsers->ferderatedUsers, etc.
It doesn't take care about these mappings anymore.
Correct?
Thanks for clarification
...Michael