This should reduce disk space usage when scraping targets containing metrics with identical names
such as `node_cpu_seconds_total`, histograms, quantiles, etc.
Expose `vm_timestamps_blocks_merged_total` and `vm_timestamps_bytes_saved_total` metrics for monitoring
the effectiveness of timestamp blocks merging.
Previously the `vm_slow_row_inserts_total` metric may be incremented multiple times for different data points per a single time series,
while only a single increment is needed when inserting the first data point for this time series.
Previously the time spent on inverted index search could exceed the configured `-search.maxQueryDuration`.
This commit stops searching in inverted index on query timeout.
This condition may occur after the following sequence of events:
1) A goroutine enters the loop body when len(addRowsConcurrencyCh) == cap(addRowsConcurrencyCh) inside Storage.searchTSIDs.
2) All the goroutines return from Storage.AddRows.
3) The goroutine from step 1 blocks on searchTSIDsCond.Wait() inside the loop body.
The goroutine remains blocked until the next call to Storage.AddRows, which calls searchTSIDsCond.Signal().
This may take indefinite time.
This is a follow-up commit after 12b16077c4 ,
which didn't reset the `tsidCache` in all the required places.
This could result in indefinite errors like:
missing metricName by metricID ...; this could be the case after unclean shutdown; deleting the metricID, so it could be re-created next time
Fix this by resetting the cache inside deleteMetricIDs function.
Previously the concurrency has been limited to GOMAXPROCS*2. This had little sense,
since every call to Storage.AddRows is bound to CPU, so the maximum ingestion bandwidth
is achieved when the number of concurrent calls to Storage.AddRows is limited to the number of CPUs,
i.e. to GOMAXPROCS.
Heavy queries could result in the lack of CPU resources for processing the current data ingestion stream.
Prevent this by delaying queries' execution until free resources are available for data ingestion.
Expose `vm_search_delays_total` metric, which may be used in for alerting when there is no enough CPU resources
for data ingestion and/or for executing heavy queries.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/291
This guarantees that the snapshot contains all the recently added data
from inmemory buffers when multiple concurrent calls to Storage.CreateSnapshot are performed.
Production workload shows that the index requires ~4Kb of RAM per active time series.
This is too much for high number of active time series, so let's delete this index.
Now the queries should fall back to the index for the current day instead of the index
for the recent hour. The query performance for the current day index should be good enough
given the 100M rows/sec scan speed per CPU core.
Issues fixed:
- Slow startup times. Now the index is loaded from cache during start.
- High memory usage related to superflouos index copies every 10 seconds.
Production load with >10M active time series showed it could
slow down VictoriaMetrics startup times and could eat
all the memory leading to OOM.
Remove inmemory inverted index for recent hours until thorough
testing on production data shows it works OK.
