chore: add more metrics for perf #66
Conversation
will-2012
force-pushed
the
add-more-metrics
branch
2 times, most recently
from
8b3e0cd to
70fa92a
Compare
will-2012
force-pushed
the
add-more-metrics
branch
from
70fa92a to
26dd775
Compare
| triehash := statedb.AccountHashes + statedb.StorageHashes // The time spent on tries hashing | ||
| trieUpdate := statedb.AccountUpdates + statedb.StorageUpdates // The time spent on tries update | ||
| trieRead := statedb.SnapshotAccountReads + statedb.AccountReads // The time spent on account read | ||
| trieRead += statedb.SnapshotStorageReads + statedb.StorageReads // The time spent on storage read | ||
| blockExecutionTimer.Update(ptime - trieRead) // The time spent on EVM processing | ||
| blockValidationTimer.Update(vtime - (triehash + trieUpdate)) // The time spent on block validation | ||
| blockExecutionTimer.Update(ptime) // The time spent on EVM processing | ||
| blockValidationTimer.Update(vtime) // The time spent on block validation |
There was a problem hiding this comment.
It seems that the definitions of blockExecutionTimer and blockValidationTimer have been re-defined. While I am not sure if this is intended. if yes, could you please explain why this change is better?
There was a problem hiding this comment.
Yes; this PR does not delete any metrics, but only adjusts the calculation method of metrics, because when analyzing performance bottlenecks during --metrics.expensive, blockExecutionTimer/blockValidationTimer/blockWriteTimer may be negative when deducting concurrency latency. This is a bit confusing. In fact, if we pay attention to the metrics of internal concurrent/async actions, we can directly configure a corresponding panel, and I think it is better to have clear semantics of blockExecutionTimer/blockValidationTimer/blockWriteTimer. :)
This is just a slight change and ultimately depends on the reviewers' suggestions.
There was a problem hiding this comment.
If without --metrics.expensive, These triehash/trieUpdate/..Commits are all equal to 0.
| fmt.Printf("loop print db stats db_metrics=\n%v\n", stats) | ||
| d.log.Info("loop print db stats", "comp_time", compTime, "write_delay_count", writeDelayCount, "write_delay_time", | ||
| writeDelayTime, "non_level0_comp_count", nonLevel0CompCount, "level0_comp_count", level0CompCount) |
There was a problem hiding this comment.
i noticed that the log level is info, will it result to log explosion
|
Close this PR since most of the metrics have been added in #71 |
Description
Add more metrics for perf.
Rationale
N/A.
Example
--metrics.expensiveenable more metrics for perf.Changes
Notable changes: