Change github objects in set to only strings

[xtuchyna]

Related Issues and Dependencies

Quick fix for recent PR #58 (PyGithub objects do not support hashing -> mentioned it here PyGithub/PyGithub#1826 ).

[pacospace]

lgtm, one question, do we have env variables to allow mi-scheduler to schedule only some specific tasks?

for example having env variables that can deactivate/activate (0 not active, 1 active) one task if we don't want to schedule everything at the same time, or if we analyze these things in different clusters and don't want to perform double work?

SCHEDULE_GH_REPO_ANALYSIS (in the future we can extend it to other SCHEDULE_GL_REPO_ANALYSIS for GitLab)

SCHEDULE_KEBECHET_ANALYSIS

wdyt?

[KPostOffice]

One small comment

[KPostOffice]

Just set the default for organizations and repositores to be [] here

[xtuchyna]

python is known to work unexpectedly here: if you use mutable as a default parameter, it can be modified later, which would affect all the other subsequent calls.

See here e.g.: https://web.archive.org/web/20200221224620/http://effbot.org/zone/default-values.htm

Therefore, for a better practice, I use None and then check, instead of mutables as default (even though they are not modified in body).

[xtuchyna]

@pacospace hmm, not sure if I understand correctly

for example having env variables that can deactivate/activate (0 not active, 1 active) one task if we don't want to schedule everything at the same time

Doesn't that do the limit setting in the openshift template (max. number of workflows allowed)?
Or do you mean something like a semaphore (that let's say one repository has a precedence over another) ?

if we analyze these things in different clusters and don't want to perform double work?

if there are two mi-scheduler cronjobs in two different clusters, setting up the env variables to avoid duplicit workflows would not work, or am i mistaken?

[pacospace]

@pacospace hmm, not sure if I understand correctly

for example having env variables that can deactivate/activate (0 not active, 1 active) one task if we don't want to schedule everything at the same time

Doesn't that do the limit setting in the openshift template (max. number of workflows allowed)?
Or do you mean something like a semaphore (that let's say one repository has a precedence over another) ?

if we analyze these things in different clusters and don't want to perform double work?

if there are two mi-scheduler cronjobs in two different clusters, setting up the env variables to avoid duplicit workflows would not work, or am i mistaken?

My comment is related to schedule specific analysis like a semaphore. Imagine for example:

cluster 1: SCHEDULE_GH_REPO_ANALYSIS=1, SCHEDULE_KEBECHET_ANALYSIS=0

cluster 2: SCHEDULE_GH_REPO_ANALYSIS=0, SCHEDULE_KEBECHET_ANALYSIS=1

See for example what we do for graph-refresh-job: https://github.com/thoth-station/graph-refresh-job/blob/d090fada6ba13ee212efc5729bf51bd3c45ba532/producer.py#L63

[xtuchyna]

@pacospace yes, that's a good idea :) Issue created

[xtuchyna]

/approve

[sesheta]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: xtuchyna

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [xtuchyna]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment