Node benchmarking utility #6198
Open
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
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #6198 +/- ##
==========================================
- Coverage 51.85% 51.74% -0.11%
==========================================
Files 639 641 +2
Lines 85508 85676 +168
==========================================
- Hits 44336 44335 -1
- Misses 38356 38524 +168
- Partials 2816 2817 +1 ☔ View full report in Codecov by Sentry. |
|
Submitting a report might be fun :
|
There was a problem hiding this comment.
Good work! I left few comments.
The PR will need an update after #6177 gets merged.
| return fmt.Sprintf(">> stage:%s duration_sec:%.1f duration_min:%.1f cpu_sec:%d", bs.stage, bs.duration.Seconds(), bs.duration.Minutes(), bs.cpuTimeNS/1000000000) | ||
| } | ||
|
|
||
| func maybeGetTotalMemory() uint64 { |
There was a problem hiding this comment.
consider moving to util/util.go
| benchCmd.Flags().IntVarP(&round, "round", "r", 0, "Specify the round number ( i.e. 7700000 )") | ||
| benchCmd.Flags().StringVarP(&relayAddress, "relay", "p", "", "Relay address to use ( i.e. r-ru.algorand-mainnet.network:4160 )") | ||
| benchCmd.Flags().StringVarP(&catchpointFile, "tar", "t", "", "Specify the catchpoint file (either .tar or .tar.gz) to process") | ||
| benchCmd.Flags().StringVarP(&reportJsonPath, "report", "j", "", "Specify the file to save the Json formatted report to") |
There was a problem hiding this comment.
Suggested change
| benchCmd.Flags().StringVarP(&reportJsonPath, "report", "j", "", "Specify the file to save the Json formatted report to") | |
| benchCmd.Flags().StringVarP(&reportJsonPath, "report", "j", "", "Specify the file to save the JSON formatted report to") |
| } | ||
|
|
||
| func GetCPU() int64 { | ||
| usage := new(syscall.Rusage) |
| addrs = []string{relayAddress} | ||
| } else { | ||
| //append relays | ||
| dnsaddrs, err := tools.ReadFromSRV(context.Background(), "algobootstrap", "tcp", networkName, "", false) |
There was a problem hiding this comment.
"algobootstrap" probably should not be here since they not obliged to have catchpoints except few most recent ones.
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.
go-algorand could use a standardized way to compare and benchmark the underlying hardware, ideally with a repeatable workload that closely matches a real scenario.
Users could compare their results online and make sure their hardware's performance is above the median so that network peak performance can grow with the number of new nodes.
The catchpointdump utility is the perfect first candidate for such a utility.
This patch adds a
benchcommand to the utility by combining both network and file restore scenarios. Download, SQLite loading and Merkle tree build can be benchmarked all in one go.It reuses some of the dependencies that are already in go.mod to get information about the hardware - at least on the Linux platform.
Results are optionally dumped to a JSON file and ready for submission to some central benchmark repository.
Examples
Simple Network, SSD and CPU test
Known catchpoint label, sourced from a random relay/archiver :
SSD and CPU test with local file
Benchmarking the disk and CPU part only using the already downloaded ledger snapshot:
Full benchmark with JSON report and hosted snapshot
A repeatable benchmark with a CloudFlare hosted catchpoint and report dump
Report file
Sample report.json:
{ "report": "a193cbc7-6e6a-732b-93cf-36f0c0589864", "stages": [ { "stage": "network", "duration_sec": 39, "cpu_time_sec": 59 }, { "stage": "database", "duration_sec": 795, "cpu_time_sec": 629 }, { "stage": "digest", "duration_sec": 363, "cpu_time_sec": 482 } ], "host": { "cores": 20, "log_cores": 20, "base_mhz": 2500, "max_mhz": 3500, "cpu_name": "13th Gen Intel(R) Core(TM) i5-13500", "cpu_vendor": "Intel", "mem_mb": 64105, "os": "linux", "uuid": "c3acdb4e-3937-a9a6-2266-d80ce615ef45" } }File can be uploaded to a 3rd pty benchmark site like:
curl -X POST https://benchmarks.nodely.io/api/report -d @report.json #{"success":true,"goto":"https://benchmarks.nodely.io/edit/a193cbc7-6e6a-732b-93cf-36f0c0589864"}