test(benchmarks): permission-policy overhead family#203
Merged
NathanFlurry merged 1 commit intoJul 2, 2026
Merged
Conversation
Measures what the permission matcher costs per hot op: each op runs twice — allow-everything vs a realistic default-deny policy (~15 fs allow globs + 5 network allow patterns) that still permits the op, so the delta is pure match cost. policyTax = policy_p50/allow_p50 is reported per op and any op over 1.2 emits a finding. Result on the current matcher: realistic policies are cheap — worst hot-op tax 1.07 (http_loopback_get), findings empty. Methodology validated with a temporary ~200-network-rule pathological policy (tcp_echo tax 2.02, finding emitted) which is documented in the README and not committed. Reuses the existing op programs (small_write, stat_storm, readdir_large, tcp_echo, http_loopback_get) via the per-op prepareVm hook; guest-only rows with explicit unsupported reasons elsewhere.
Member
Author
|
Stack for rivet-dev/secure-exec
Get stack: |
railway-app
Bot
temporarily deployed
to
secure-exec / secure-exec-pr-203
NathanFlurry
deleted the
stack/test-benchmarks-permission-policy-overhead-family-qwowowmr
branch
railway-app
Bot
requested a deployment
to
rivet-frontend / rivet-pr-5343
railway-app
Bot
requested a deployment
to
rivet-frontend / rivet-pr-5314
railway-app
Bot
requested a deployment
to
rivet-frontend / rivet-pr-5328
railway-app
Bot
requested a deployment
to
rivet-frontend / agentos-pr-1507
railway-app
Bot
requested a deployment
to
rivet-frontend / rivet-pr-5333
railway-app
Bot
requested a deployment
to
rivet-frontend / agentos-pr-1510
railway-app
Bot
requested a deployment
to
rivet-frontend / rivet-pr-5324
railway-app
Bot
requested a deployment
to
rivet-frontend / rivet-pr-5338
railway-app
Bot
requested a deployment
to
rivet-frontend / rivet-pr-5320
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.
Measures what the permission matcher costs per hot op: each op runs twice —
allow-everything vs a realistic default-deny policy (~15 fs allow globs + 5
network allow patterns) that still permits the op, so the delta is pure match
cost. policyTax = policy_p50/allow_p50 is reported per op and any op over 1.2
emits a finding.
Result on the current matcher: realistic policies are cheap — worst hot-op
tax 1.07 (http_loopback_get), findings empty. Methodology validated with a
temporary ~200-network-rule pathological policy (tcp_echo tax 2.02, finding
emitted) which is documented in the README and not committed.
Reuses the existing op programs (small_write, stat_storm, readdir_large,
tcp_echo, http_loopback_get) via the per-op prepareVm hook; guest-only rows
with explicit unsupported reasons elsewhere.