Make stdout maxBuffer length to be configurable

[vrubezhny]

The default value is set to 4Gb.

Fixes: #3925

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 45.04%. Comparing base (da60441) to head (7570b24).
Report is 170 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #4061       +/-   ##
===========================================
+ Coverage   32.37%   45.04%   +12.66%     
===========================================
  Files          85       85               
  Lines        6505     6720      +215     
  Branches     1349     1404       +55     
===========================================
+ Hits         2106     3027      +921     
+ Misses       4399     3693      -706     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[datho7561]

With this change, I am able to open the "Pods" and "Deployments" sections if I create 100 deployments. I am going to stress test 400 to see what happens.

[vrubezhny]

With this change, I am able to open the "Pods" and "Deployments" sections if I create 100 deployments. I am going to stress test 400 to see what happens.

The number of pods really matters: the error (ERR_CHILD_PROCESS_STDIO_MAXBUFFER) happens when executing oc get pods -o=json , and according to the reporter and @lgrossma 's experience it takes place when the number of pods is 200+ (with old default stdout buffer limit equal to 2MiB. If you're using this PR when testing you can set openshiftToolkit.execMaxBufferLength to 2 ( === 2*1024*1024) to make it word exactly as it did before the change) .

[datho7561]

4 MB allows for viewing over 400 deployments, this should (hopefully) be plenty

[datho7561]

I used

#!/usr/bin/lua

for i = 1, 400, 1 do
  os.execute(string.format('oc create deployment --image quay.io/davthomp/pet-alpine my-pet-deploy%d', i));
end

To create 400 deployments, each with one pod. I opened both the deployments and the pods sections. No crashes occured, it loaded properly.