It has been appeared that some production workloads could suffer for some time
after every reset of the previous cache when it gets less than 5% of requests
after the needed item isn't found in the current cache. This could result
in reduced cache hit rates, which, in turn, could increase CPU, disk IO and RAM
usage needed for reading, unpacking and caching the missed data from disk.
This commit reduces the cache miss threshold for resetting the previous cache from 5% to 1%.
This should reduce the possible negative impact after each cache reset by at least 5x,
while reducing the total memory used by caches.
This is a follow-up for d906d8573e
The issue was in the `labels := dst[offset:]` line in the beginning of appendExtraLabels() function.
The `dst` may be re-allocated when adding extra labels to it. In this case the addition of `exported_`
prefix to labels inside `labels` slice become invisible in the returned `dst` labels.
While at it, properly handle some corner cases:
- Add additional `exported_` prefix to clashing metric labels with already existing `exported_` prefix.
- Store scraped metric names in `exported___name__` label if scrape target contains `__name__` label.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/3278
Thanks to @jplanckeel for the initial attempt to fix this issue
at https://github.com/VictoriaMetrics/VictoriaMetrics/pull/3281
Previously netstorage.MustStop() call didn't free up all the resources,
so the subsequent call to nestorage.Init() would panic.
This allows writing tests, which call nestorage.Init() + nestorage.MustStop() in a loop.
Blocks outside the configured retention are eventually deleted during background merge.
But such blocks may reside in the storage for long time until background merge.
Previously VictoriaMetrics could spend additional CPU time on processing such blocks
during search queries. Now these blocks are skipped.
The searchTSIDs function was searching for metricIDs matching the the given tag filters
and then was locating the corresponding TSID entries for the found metricIDs.
The TSID entries aren't needed when searching for time series names (aka MetricName),
so this commit removes the uneeded TSID search from the implementation of /api/v1/series API.
This improves perfromance of /api/v1/series calls.
This commit also improves performance a bit for /api/v1/query and /api/v1/query_range calls,
since now these calls cache small metricIDs instead of big TSID entries
in the indexdb/tagFilters cache (now this cache is named indexdb/tagFiltersToMetricIDs)
without the need to compress the saved entries in order to save cache space.
This commit also removes concurrency limiter during searching for matching time series,
which was introduced in 8f16388428, since the concurrency
for all the read queries is already limited with -search.maxConcurrentRequests command-line flag.
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/issues/648
This increases the maximum time for cache population with new entries from 20 minutes to 40 minutes.
This
This change shouldn't increase memory usage for caches, since the prev cache cleaner
should free up memory by deleting unused prev cache as soon as possible.
See 08ca45d238 for details on prev cache cleaner.
The panic has been introduced in 68f3a02589
While at it, add padding to shard structs in order to avoid false sharing on mordern CPUs
This should improve scalability on systems with many CPU cores
This means that the majority of requests are successfully served from the current cache,
so the previous cache can be reset in order to free up memory.
