Clear ApplicationLogEvent and SkaffoldLogEvent buffers on new dev iteration

[MarlonGamez]

Related: #5368 #5370

Description
This change resets the buffers containing SkaffoldLogEvents and ApplicationLogEvents on a new dev iteration, to prevent increased RAM usage over time.

[codecov]

Codecov Report

Merging #5993 (8f14bd0) into master (4afe45b) will decrease coverage by 0.09%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #5993      +/-   ##
==========================================
- Coverage   70.76%   70.66%   -0.10%     
==========================================
  Files         460      462       +2     
  Lines       17746    17804      +58     
==========================================
+ Hits        12558    12582      +24     
- Misses       4260     4289      +29     
- Partials      928      933       +5     

Impacted Files	Coverage Δ
pkg/skaffold/event/v2/event.go	76.38% <100.00%> (+0.22%)	⬆️
pkg/skaffold/deploy/deploy_mux.go	75.67% <0.00%> (-9.18%)	⬇️
pkg/skaffold/util/tar.go	50.66% <0.00%> (-5.34%)	⬇️
pkg/skaffold/kubernetes/logger/log.go	37.50% <0.00%> (-4.71%)	⬇️
pkg/skaffold/kubernetes/watcher.go	80.48% <0.00%> (-3.89%)	⬇️
pkg/skaffold/docker/image.go	78.13% <0.00%> (-1.40%)	⬇️
pkg/skaffold/runner/v1/deploy.go	53.00% <0.00%> (-0.47%)	⬇️
pkg/skaffold/runner/v1/dev.go	72.98% <0.00%> (-0.13%)	⬇️
pkg/skaffold/deploy/kpt/kpt.go	81.75% <0.00%> (-0.10%)	⬇️
pkg/skaffold/deploy/helm/deploy.go	79.16% <0.00%> (-0.08%)	⬇️
... and 8 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4afe45b...8f14bd0. Read the comment docs.

[tejal29]

should we allocate new memory instead of nil to make sure there is no nil point exception

[MarlonGamez]

@tejal29 I'm unsure, since there are no fields of a slice that we can access so I don't think that could happen. I think that functionally, a nil and empty slice would behave the same. Iterating over it just results in zero iterations, like an empty slice.
This way is just cleaner to me over writing out a slice literal. I'm happy to change if you prefer tho!

[tejal29]

@tejal29 I'm unsure, since there are no fields of a slice that we can access so I don't think that could happen. I think that functionally, a nil and empty slice would behave the same. Iterating over it just results in zero iterations, like an empty slice.
This way is just cleaner to me over writing out a slice literal. I'm happy to change if you prefer tho!

ah. I guess iterating would be fine. I had looked up usage last time and looks like we pass the pointer to these fields.

return ev.forEach(&ev.applicationLogListeners, &ev.applicationLogs, &ev.applicationLogsLock, callback)

My concern is this might end up into nil pointer exception?

[MarlonGamez]

@tejal29 made the change to use slice literals 👍🏼