[DRAFT] loadgen: add probabilistic memo distribution across the workflow tree#420
Draft
tdeebswihart wants to merge 5 commits into
Draft
[DRAFT] loadgen: add probabilistic memo distribution across the workflow tree#420tdeebswihart wants to merge 5 commits into
tdeebswihart wants to merge 5 commits into
Conversation
…ghput_stress Adds a --option payload-distribution-json knob (with @file support) that drives activity payload sizes from the existing DistributionField framework (discrete, zipf, uniform, normal, fixed), replacing hardcoded 256-byte payloads in throughput_stress. Payload activity input bytes are now incompressible pseudo-random (deterministic per size, replay-safe) so payloads occupy their full configured size in history. Also fixes a latent data race in discreteDistribution's lazy cache init (sync.Once).
tdeebswihart
changed the title
tdeebswihart
force-pushed
the
tim/memo-payloadsize
branch
from
July 14, 2026 14:27
6d3d765 to
070b0e1
Compare
Adds a --option memo-distribution-json knob (probability + size distribution,
capped at 40 KB) that attaches an incompressible memo blob ("MemoBlob") to
throughput_stress workflows: the root, child workflows, and continue-as-new
runs. Probability is rolled independently per node from a memo-dedicated rng
(does not perturb payload sampling); along a continue-as-new chain the memo is
sticky/additive, carried forward client-side so behavior is identical across
all worker SDK languages. Updates all six language workers (go, python,
typescript, java, dotnet, ruby) to honor the child-workflow and
continue-as-new Memo proto fields, passing the pre-encoded payload through each
SDK's raw/passthrough path to avoid double-encoding.
tdeebswihart
force-pushed
the
tim/memo-payloadsize
branch
from
July 14, 2026 16:19
1f72dba to
80165ec
Compare
tdeebswihart
force-pushed
the
tim/tputstress-payloadsize
branch
5 times, most recently
from
July 14, 2026 19:30
f64e55d to
b94b83d
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
What was changed
Stacked on #419.
Adds a
--option memo-distribution-jsonknob (probability + size distribution, capped at 40 KB) that attaches an incompressible memo blob (MemoBlob) tothroughput_stressworkflows throughout the tree: the root, child workflows, and continue-as-new runs.Memoproto fields, passing the pre-encoded payload through each SDK's raw/passthrough path to avoid double-encoding.Why?
Production workflow records carry memos that grow mutable state. Attaching probabilistic, distribution-sized memos across the tree makes throughput_stress records more production-like.
How was this tested
go build,py_compile,gradlew compileJava,tsc,dotnet build,ruby -c).binary/plainuniformly (no double-encoding), with no non-determinism.