debugger_v2: start multiplexer lazily #4659
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
The DebuggerV2 plugin reads data outside of the usual summary-based
workflows, and has its own loading infrastructure for doing so. Until
this patch, the plugin would start recursively scanning the user’s log
directory at plugin initialization time—even before the server launches.
This can take a lot of resources and introduce a lot of contention,
especially on remote file systems.
This patch defers all such logic until users of the frontend explicitly
activate the plugin. Startup time improves greatly (see below).
Test Plan:
When pointing at
gs://tensorboard-bench-logs, the time between processstart and the “server ready” line drops from 9.5 seconds to 1.7 seconds,
as measured by:
After generating test data as described in the tutorial, the plugin is
indeed inactive by default, and does indeed still appear to work if
manually activated.
wchargin-branch: dbgv2-lazy-load