Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[system-probe] Add packet counting to existing UDP byte counters #28546

Merged
merged 10 commits into from
Sep 5, 2024

Conversation

pimlu
Copy link
Contributor

@pimlu pimlu commented Aug 19, 2024

What does this PR do?

This PR begins tracing UDP packet counts in the agent and adds some tests to ensure it works as expected for both IPv4 and IPv6.

Motivation

Context - NPM-2973. The agent already has tracked UDP byte counts so it was determined we can leverage the existing kprobes to track packet counts in the same places.

Additional Notes

Possible Drawbacks / Trade-offs

This does not track UDP sends that have corking or MSG_MORE which are not commonly used. Doing this requires tracing udp_send_skb which probably will have to wait until after we deprecate prebuilt as mentioned in this thread

Describe how to test/QA your changes

This PR leverages the existing test suite and adds some new ones:

DD_SYSTEM_PROBE_CONFIG_BPF_DEBUG=true sudo -E go test -v -tags "linux linux_bpf npm test" ./pkg/network/tracer --run TestTracerSuite

@pr-commenter
Copy link

pr-commenter bot commented Aug 19, 2024

Test changes on VM

Use this command from test-infra-definitions to manually test this PR changes on a VM:

inv create-vm --pipeline-id=42853416 --os-family=ubuntu

Note: This applies to commit b9a0ad6

@pr-commenter
Copy link

pr-commenter bot commented Aug 19, 2024

eBPF complexity changes

Summary result: ✅ - stable

  • Highest complexity change (%): +1.55%
  • Highest complexity change (abs.): +16 instructions
  • Programs that were above the 85.0% limit of instructions and are now below: 0
  • Programs that were below the 85.0% limit of instructions and are now above: 0
tracer details

tracer

Program Avg. complexity Distro with highest complexity Distro with lowest complexity
kretprobe__ip_make_skb 🔴 1148.7 (+16.0, +1.41%) fedora_38/arm64: 🔴 1303.0 (+16.0, +1.24%) centos_8/arm64: 🔴 1048.0 (+16.0, +1.55%)
kretprobe__tcp_close_clean_protocols ⚪ 212.4 (+0.0, +0.00%) centos_8/arm64: ⚪ 225.0 (+0.0, +0.00%) debian_10/x86_64: ⚪ 199.0 (+0.0, +0.00%)
kretprobe__tcp_close_flush ⚪ 216.4 (+0.0, +0.00%) ubuntu_18.04/x86_64: ⚪ 217.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 216.0 (+0.0, +0.00%)
kretprobe__tcp_done_flush ⚪ 216.4 (+0.0, +0.00%) ubuntu_18.04/x86_64: ⚪ 217.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 216.0 (+0.0, +0.00%)
kretprobe__tcp_read_sock ⚪ 1036.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 1128.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 990.0 (+0.0, +0.00%)
kretprobe__tcp_recvmsg ⚪ 1036.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 1128.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 990.0 (+0.0, +0.00%)
kretprobe__tcp_retransmit_skb ⚪ 475.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 559.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 434.0 (+0.0, +0.00%)
kretprobe__tcp_sendmsg ⚪ 1036.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 1128.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 990.0 (+0.0, +0.00%)
kretprobe__tcp_sendpage ⚪ 1036.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 1128.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 990.0 (+0.0, +0.00%)
kretprobe__udp_destroy_sock ⚪ 217.4 (+0.0, +0.00%) ubuntu_18.04/x86_64: ⚪ 218.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 217.0 (+0.0, +0.00%)
kretprobe__udp_recvmsg ⚪ 9.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 9.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 9.0 (+0.0, +0.00%)
kretprobe__udp_recvmsg_pre_4_7_0 ⚪ 1386.0 (+0.0, +0.00%) fedora_38/arm64: ⚪ 1612.0 (+0.0, +0.00%) amazon_5.4/x86_64: ⚪ 1129.0 (+0.0, +0.00%)
kretprobe__udp_sendpage 🔴 894.0 (+8.0, +0.90%) fedora_38/arm64: 🔴 965.0 (+8.0, +0.84%) centos_8/arm64: 🔴 847.0 (+8.0, +0.95%)
kretprobe__udpv6_destroy_sock ⚪ 217.4 (+0.0, +0.00%) ubuntu_18.04/x86_64: ⚪ 218.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 217.0 (+0.0, +0.00%)
kretprobe__udpv6_recvmsg ⚪ 9.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 9.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 9.0 (+0.0, +0.00%)
kretprobe__udpv6_recvmsg_pre_4_7_0 ⚪ 1386.0 (+0.0, +0.00%) fedora_38/arm64: ⚪ 1612.0 (+0.0, +0.00%) amazon_5.4/x86_64: ⚪ 1129.0 (+0.0, +0.00%)
socket__classifier_dbs ⚪ 2427.0 (+0.0, +0.00%) fedora_38/arm64: ⚪ 2434.0 (+0.0, +0.00%) amazon_5.4/x86_64: ⚪ 2420.0 (+0.0, +0.00%)
socket__classifier_entry ⚪ 2547.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 2562.0 (+0.0, +0.00%) amazon_5.4/x86_64: ⚪ 2534.0 (+0.0, +0.00%)
socket__classifier_grpc ⚪ 6754.0 (+0.0, +0.00%) fedora_38/arm64: ⚪ 6775.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 6719.0 (+0.0, +0.00%)
socket__classifier_queues ⚪ 7023.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 8247.0 (+0.0, +0.00%) amazon_5.4/x86_64: ⚪ 4617.0 (+0.0, +0.00%)
tracepoint__net__net_dev_queue ⚪ 971.7 (+0.0, +0.00%) fedora_38/arm64: ⚪ 1183.0 (+0.0, +0.00%) centos_8/arm64: ⚪ 800.0 (+0.0, +0.00%)

This report was generated based on the complexity data for the current branch HEAD (pipeline [https://gitlab.ddbuild.io/DataDog/datadog-agent/-/pipelines/42483018]42483018) and the main branch (commit fe13dea). Objects without changes are not reported. Contact #ebpf-platform if you have any questions/feedback.

Table complexity legend: 🔵 - new; ⚪ - unchanged; 🟢 - reduced; 🔴 - increased

@akarpz
Copy link
Contributor

akarpz commented Aug 19, 2024

can you add a description? also it's helpful to put the ticket # in the title and/or under the motivation section

@pr-commenter
Copy link

pr-commenter bot commented Aug 21, 2024

Regression Detector

Regression Detector Results

Run ID: f71704c0-7619-46c5-8e50-c37c9f99ecf3 Metrics dashboard Target profiles

Baseline: 207b216
Comparison: b9a0ad6

Performance changes are noted in the perf column of each table:

  • ✅ = significantly better comparison variant performance
  • ❌ = significantly worse comparison variant performance
  • ➖ = no significant change in performance

No significant changes in experiment optimization goals

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

There were no significant changes in experiment optimization goals at this confidence level and effect size tolerance.

Fine details of change detection per experiment

perf experiment goal Δ mean % Δ mean % CI links
basic_py_check % cpu utilization +0.96 [-1.59, +3.51] Logs
otel_to_otel_logs ingress throughput +0.28 [-0.53, +1.09] Logs
uds_dogstatsd_to_api ingress throughput -0.00 [-0.00, +0.00] Logs
tcp_dd_logs_filter_exclude ingress throughput -0.00 [-0.01, +0.01] Logs
file_tree memory utilization -0.21 [-0.27, -0.15] Logs
idle memory utilization -0.53 [-0.57, -0.50] Logs
uds_dogstatsd_to_api_cpu % cpu utilization -0.73 [-1.51, +0.05] Logs
pycheck_lots_of_tags % cpu utilization -0.82 [-3.18, +1.54] Logs
tcp_syslog_to_blackhole ingress throughput -4.32 [-16.70, +8.06] Logs

Bounds Checks

perf experiment bounds_check_name replicates_passed
idle memory_usage 10/10

Explanation

A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".

For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:

  1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

  2. Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.

  3. Its configuration does not mark it "erratic".

@pimlu pimlu force-pushed the stuart.geipel/udp-counting branch from 9ba64e0 to ebe8f63 Compare August 22, 2024 14:54
@pimlu pimlu force-pushed the stuart.geipel/udp-counting branch from 9b93718 to 036bfaa Compare August 26, 2024 18:25
@pimlu pimlu requested a review from a team as a code owner August 26, 2024 18:35
Copy link
Contributor

@hestonhoffman hestonhoffman left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Docs 👍

@pimlu
Copy link
Contributor Author

pimlu commented Sep 5, 2024

/merge

@dd-devflow
Copy link

dd-devflow bot commented Sep 5, 2024

🚂 MergeQueue: pull request added to the queue

The median merge time in main is 23m.

Use /merge -c to cancel this operation!

@dd-mergequeue dd-mergequeue bot merged commit d29f884 into main Sep 5, 2024
295 checks passed
@dd-mergequeue dd-mergequeue bot deleted the stuart.geipel/udp-counting branch September 5, 2024 18:02
@github-actions github-actions bot added this to the 7.58.0 milestone Sep 5, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants