1308782 - Don't need to download jobtype and jobgroup information on load

Status: RESOLVED FIXED

(Tree Management :: Treeherder, defect)

Description

[William Lachance (:wlach)]

We currently download the jobtype and jobgroup information every time we load treeherder to establish a "sorting order" for this information. 

This is rather slow in the case of job type information (3 seconds from my local wifi network) and in point of fact we don't need to do this -- we can just sort the job and group information dynamically using the same algorithm and have the same result.

With this enabled, treeherder is a good chunk faster to load initially, most especially on things like try pushes where downloading the jobtype information was actually the bottleneck!

Comment 1

[GitHub Autolander Bot]

Attachment: [treeherder] wlach:1308782 > mozilla:master

Comment 2

[William Lachance (:wlach)]

Comment on attachment 8799247 [details] [review]
[treeherder] wlach:1308782 > mozilla:master

Noticed this while working on some other stuff. This isn't urgent, but does provide a nice speed-up when loading treeherder.

Comment 3

[Cameron Dawson [:camd]]

Comment on attachment 8799247 [details] [review]
[treeherder] wlach:1308782 > mozilla:master

Hey Will--  this is great!  Just a couple of minor cleanup things here and there.  :)

Comment 4

[William Lachance (:wlach)]

Comment on attachment 8799247 [details] [review]
[treeherder] wlach:1308782 > mozilla:master

Made the changes you requested, could you have a quick look?

Comment 5

[William Lachance (:wlach)]

Comment on attachment 8799247 [details] [review]
[treeherder] wlach:1308782 > mozilla:master

Gah, I meant to r? camd, not myself. :)

Comment 6

[Cameron Dawson [:camd]]

Comment on attachment 8799247 [details] [review]
[treeherder] wlach:1308782 > mozilla:master

Awesome!  :)

Comment 7

[Treeherder GitHub Bugbot]

Commit pushed to master at https://github.com/mozilla/treeherder

https://github.com/mozilla/treeherder/commit/700ef6bec2b681fdc564de25ada1f3a954cba521
Bug 1308782 - Don't download job and group data on load (#1912)

Instead, just calculate on the fly with the job information we have.
This makes the display of the first job data about 3 seconds faster in
my preliminary testing.