feat: add TunnelTime metric #171
Merged
Conversation
This file contains 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
…per access key and location.
|
Actually just realized I need to still put this behind a flag, so reverting to draft. But feel free to leave comments on what's there already. |
fortuna
requested changes
Mar 19, 2024
`MustParseAddr` is meant to be used for tests only, according to the docs.
fortuna
requested changes
Mar 21, 2024
Instead look it up when we first create an `activeClient`.
sbruens
force-pushed
the
sbruens/ip-key-metrics
branch
from
188ca9d
to
ced36bd
Compare
|
Made some big changes to use the Prometheus Collector interface (thanks for the pointer, much better). |
fortuna
requested changes
Mar 25, 2024
There was a problem hiding this comment.
Thanks, this looks much cleaner. I think the service code looks pretty good, only one small note.
As for the Prometheus code, I'm pretty sure you need to accumulate the counter. Prometheus only mirrors it. So you will need to keep track of the two counters (per key and per location+asn).
There was a problem hiding this comment.
It would be helpful to run the benchmarks with and without your changes, then diff, to make sure that's no significant difference we are not accounting for.
There was a problem hiding this comment.
panic: duplicate metrics collector registration attempted [recovered]
goos: linux
goarch: amd64
pkg: github.com/Jigsaw-Code/outline-ss-server/internal/integration_test
cpu: Intel(R) Xeon(R) W-2135 CPU @ 3.70GHz
│ bench_master.log │ bench_branch.log │
│ sec/op │ sec/op vs base │
TCPThroughput-6 11.51µ ± 2% 11.54µ ± 3% ~ (p=0.796 n=10)
TCPMultiplexing-6 122.2µ ± 9% 125.2µ ± 8% ~ (p=0.190 n=10)
UDPEcho-6 135.4µ ± 53% 128.9µ ± 35% ~ (p=0.853 n=10)
UDPManyKeys-6 1.449m ± 66% 1.113m ± 78% ~ (p=0.165 n=10)
geomean 128.9µ 120.0µ -6.93%
│ bench_master.log │ bench_branch.log │
│ mbps │ mbps vs base │
TCPThroughput-6 694.8 ± 2% 693.0 ± 2% ~ (p=0.781 n=10)
│ bench_master.log │ bench_branch.log │
│ B/op │ B/op vs base │
TCPThroughput-6 11.00 ± 55% 10.00 ± 50% ~ (p=0.897 n=10)
TCPMultiplexing-6 81.96Ki ± 3% 82.05Ki ± 2% ~ (p=0.684 n=10)
UDPEcho-6 5.282Ki ± 1% 5.287Ki ± 1% ~ (p=0.540 n=10)
UDPManyKeys-6 128.3Ki ± 63% 128.3Ki ± 74% ~ (p=0.382 n=10)
geomean 4.942Ki 4.828Ki -2.30%
│ bench_master.log │ bench_branch.log │
│ allocs/op │ allocs/op vs base │
TCPThroughput-6 0.000 ± 0% 0.000 ± 0% ~ (p=1.000 n=10) ¹
TCPMultiplexing-6 646.5 ± 5% 644.0 ± 3% ~ (p=0.493 n=10)
UDPEcho-6 97.00 ± 0% 97.00 ± 0% ~ (p=1.000 n=10) ¹
UDPManyKeys-6 909.0 ± 58% 909.0 ± 69% ~ (p=0.960 n=10)
geomean ² -0.10% ²
¹ all samples are equal
² summaries must be >0 to compute geomean
pkg: github.com/Jigsaw-Code/outline-ss-server/internal/slicepool
│ bench_master.log │ bench_branch.log │
│ sec/op │ sec/op vs base │
Pool-6 49.55n ± 26% 43.32n ± 43% ~ (p=0.645 n=10)
│ bench_master.log │ bench_branch.log │
│ B/op │ B/op vs base │
Pool-6 0.000 ± 0% 0.000 ± 0% ~ (p=1.000 n=10) ¹
¹ all samples are equal
│ bench_master.log │ bench_branch.log │
│ allocs/op │ allocs/op vs base │
Pool-6 0.000 ± 0% 0.000 ± 0% ~ (p=1.000 n=10) ¹
¹ all samples are equal
pkg: github.com/Jigsaw-Code/outline-ss-server/ipinfo
│ bench_master.log │ bench_branch.log │
│ sec/op │ sec/op vs base │
GetIPInfoFromAddr-6 748.1n ± 38% 700.5n ± 29% ~ (p=0.515 n=10)
NewMMDBIPInfoMap-6 19.87µ ± 36% 19.15µ ± 28% ~ (p=0.912 n=10)
geomean 3.856µ 3.663µ -5.01%
│ bench_master.log │ bench_branch.log │
│ B/op │ B/op vs base │
GetIPInfoFromAddr-6 56.00 ± 0% 56.00 ± 0% ~ (p=1.000 n=10) ¹
NewMMDBIPInfoMap-6 1.617Ki ± 0% 1.617Ki ± 0% ~ (p=1.000 n=10) ¹
geomean 304.5 304.5 +0.00%
¹ all samples are equal
│ bench_master.log │ bench_branch.log │
│ allocs/op │ allocs/op vs base │
GetIPInfoFromAddr-6 4.000 ± 0% 4.000 ± 0% ~ (p=1.000 n=10) ¹
NewMMDBIPInfoMap-6 61.00 ± 0% 61.00 ± 0% ~ (p=1.000 n=10) ¹
geomean 15.62 15.62 +0.00%
¹ all samples are equal
pkg: github.com/Jigsaw-Code/outline-ss-server/service
│ bench_master.log │ bench_branch.log │
│ sec/op │ sec/op vs base │
Locking-6 247.1n ± 40% 190.8n ± 96% ~ (p=0.684 n=10)
Snapshot-6 2.927µ ± 18% 2.717µ ± 12% ~ (p=0.247 n=10)
ReplayCache_Creation-6 24.60µ ± 8% 24.66µ ± 5% ~ (p=0.912 n=10)
ReplayCache_Max-6 111.2n ± 10% 102.3n ± 17% ~ (p=0.516 n=10)
ReplayCache_Min-6 294.8n ± 86% 306.0n ± 55% ~ (p=1.000 n=10)
ReplayCache_Parallel-6 262.8n ± 10% 242.8n ± 14% ~ (p=0.247 n=10)
ServerSaltGenerator-6 1.903µ ± 5% 1.840µ ± 3% ~ (p=0.052 n=10)
TCPFindCipherFail-6 932.6µ ± 128% 1039.9µ ± 25% ~ (p=0.631 n=10)
TCPFindCipherRepeat-6 121.48µ ± 736% 80.13µ ± 1178% ~ (p=0.315 n=10)
UDPUnpackFail-6 1.063m ± 46% 1.201m ± 27% ~ (p=0.684 n=10)
UDPUnpackRepeat-6 312.22µ ± 89% 62.81µ ± 1226% ~ (p=0.971 n=10)
UDPUnpackSharedKey-6 13.29µ ± 25% 13.10µ ± 20% ~ (p=0.853 n=10)
geomean 8.517µ 7.036µ -17.39%
│ bench_master.log │ bench_branch.log │
│ B/op │ B/op vs base │
Locking-6 8.000 ± 0% 8.000 ± 0% ~ (p=1.000 n=10) ¹
Snapshot-6 8.000Ki ± 0% 8.000Ki ± 0% ~ (p=1.000 n=10) ¹
ReplayCache_Creation-6 208.0Ki ± 0% 208.0Ki ± 0% ~ (p=1.000 n=10)
ReplayCache_Max-6 10.00 ± 0% 10.00 ± 0% ~ (p=1.000 n=10) ¹
ReplayCache_Min-6 95.00 ± 0% 95.00 ± 0% ~ (p=1.000 n=10) ¹
ReplayCache_Parallel-6 10.00 ± 0% 10.00 ± 0% ~ (p=1.000 n=10) ¹
ServerSaltGenerator-6 976.0 ± 0% 976.0 ± 0% ~ (p=1.000 n=10) ¹
TCPFindCipherFail-6 116.9Ki ± 0% 116.9Ki ± 0% ~ (p=0.724 n=10)
TCPFindCipherRepeat-6 27.37Ki ± 188% 25.64Ki ± 207% ~ (p=0.353 n=10)
UDPUnpackFail-6 116.7Ki ± 0% 116.7Ki ± 0% ~ (p=1.000 n=10) ¹
UDPUnpackRepeat-6 19.603Ki ± 82% 4.918Ki ± 805% ~ (p=1.000 n=10)
UDPUnpackSharedKey-6 1.166Ki ± 0% 1.166Ki ± 0% ~ (p=1.000 n=10) ¹
geomean 1.768Ki 1.567Ki -11.37%
¹ all samples are equal
│ bench_master.log │ bench_branch.log │
│ allocs/op │ allocs/op vs base │
Locking-6 1.000 ± 0% 1.000 ± 0% ~ (p=1.000 n=10) ¹
Snapshot-6 1.000 ± 0% 1.000 ± 0% ~ (p=1.000 n=10) ¹
ReplayCache_Creation-6 2.000 ± 0% 2.000 ± 0% ~ (p=1.000 n=10) ¹
ReplayCache_Max-6 0.000 ± 0% 0.000 ± 0% ~ (p=1.000 n=10) ¹
ReplayCache_Min-6 1.000 ± 0% 1.000 ± 0% ~ (p=1.000 n=10) ¹
ReplayCache_Parallel-6 0.000 ± 0% 0.000 ± 0% ~ (p=1.000 n=10) ¹
ServerSaltGenerator-6 10.00 ± 0% 10.00 ± 0% ~ (p=1.000 n=10) ¹
TCPFindCipherFail-6 1.605k ± 0% 1.606k ± 0% ~ (p=0.771 n=10)
TCPFindCipherRepeat-6 123.00 ± 579% 99.50 ± 738% ~ (p=0.403 n=10)
UDPUnpackFail-6 1.602k ± 0% 1.602k ± 0% ~ (p=1.000 n=10) ¹
UDPUnpackRepeat-6 259.00 ± 86% 56.50 ± 967% ~ (p=1.000 n=10)
UDPUnpackSharedKey-6 17.00 ± 0% 17.00 ± 0% ~ (p=1.000 n=10) ¹
geomean ² -13.46% ²
¹ all samples are equal
² summaries must be >0 to compute geomean
fortuna
reviewed
Mar 25, 2024
fortuna
approved these changes
Mar 26, 2024
fortuna
reviewed
Mar 26, 2024
Commit suggestions. Co-authored-by: Vinicius Fortuna <[email protected]>
fortuna
approved these changes
Mar 26, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Over a given timeframe, we observe the duration that a given IPKey pair is sending traffic through the server (i.e. how long is it active) and measure it, aggregated by key or location. This will allow service providers to measure TunnelTime.