Re: Whats latest on Glusto + GD2 integration?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Thank you Rahul for the report. This does help to keep community up to date on the effort being put up here and understand where the things stand. Some comments inline.

On Sun, Nov 4, 2018 at 8:01 PM Rahul Hinduja <rhinduja@xxxxxxxxxx> wrote:
Hello,

Over past few weeks, few folks are engaged in integrating gd2 with existing glusto infrastructure/cases. This email is an attempt to provide the high level view of the work that's done so far and next.

Whats Done.
  • Libraries incorporated / under review: 
    • Gluster Base Class and setup.py file required to read config file and install all the packages
    • Exception and lib-utils file required for all basic test cases
    • Common rest methods(Post, Get,  Delete), to handle rest api’s
    • Peer management libraries
    • Basic Volume management libraries
    • Basic Snapshot libraries
    • Self-heal libraries
    • Glusterd init
    • Mount operations
    • Device operations
Note: I request you all to provide review comments on the libraries that are submitted. Over this week, Akarsha and Vaibhavi will try to get the review comments incorporated and to get these libraries to closure.
  • Where is the repo?
  • Are we able to consume gd1 cases into gd2?
    • We tried POC to run glusterd and snapshot test cases (one-by-one) via modified automation and libraries. Following are the highlights:
      • We were able to run 20 gd1 cases out of which 8 passed and 12 failed.
      • We were able to run 11 snapshot cases out of which 7 passed and 4 failed.
    • Reason for failures:
      • Because of different volume options with gd1/gd2
Just to clarify here, we have an open GD2 issue  https://github.com/gluster/glusterd2/issues/739 which is being worked on and that should help us to achieve this backward compatibility.
      • Due to different error or output format between gd1/gd2

We need to move towards parsing error codes than the error messages. I'm aware that with GD1/CLI such infra was missing, but now that GD2 offers specific error codes, all command failures need to be parsed through error/ret codes in GD2. I believe the library/tests need to be modified accordingly to cater to this need to handle both GD1/GD2 based failures.

What's next?
  • We have identified few gaps when we triggered glusterd and snapshot cases. Details in column C of  [2]. We are in the process of closing those gaps so that we don't have to hard-code or skip any functions in the test cases.
  • Develop additional/Modify existing libraries for the cases which got skipped.
  • Need to check on the volume options and error message or output format. This is being brought up in gd2 standup to freeze on the parity and rework at functional code level or automation code level.
  • I am aiming to provide the bi-weekly report on this integration work to the mailing list

For more information/collaboration, please reach-out to:

Regards,

Rahul Hinduja
_______________________________________________
Gluster-devel mailing list
Gluster-devel@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-devel
_______________________________________________
Gluster-devel mailing list
Gluster-devel@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-devel

[Index of Archives]     [Gluster Users]     [Ceph Users]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux