Table of Contents

Query Frontend #

The thanos query-frontend command implements a service that can be put in front of Thanos Queriers to improve the read path. It is based on the Cortex Query Frontend component so you can find some common features like Splitting and Results Caching.

Query Frontend is fully stateless and horizontally scalable.

Example command to run Query Frontend:

thanos query-frontend \
    --http-address     "" \

_NOTE: Currently only range queries (/api/v1/query_range API call) are actually processed through Query Frontend. All other API calls just directly go to the downstream Querier, which means only range queries are split and cached. But we are planning to support instant queries as well.

For more information please check out initial design proposal.

Features #

Splitting #

Query Frontend splits a long query into multiple short queries based on the configured --query-range.split-interval flag. The default value of --query-range.split-interval is 24h. When caching is enabled it should be greater than 0.

There are some benefits from query splitting:

  1. Safe guard. It prevents large queries from causing OOM issues to Queries.
  2. Better parallelization.
  3. Better load balancing for Queries.

Retry #

Query Frontend supports a retry mechanism to retry query when HTTP requests are failing. There is a --query-range.max-retries-per-request flag to limit the maximum retry times.

Caching #

Query Frontend supports caching query results and reuses them on subsequent queries. If the cached results are incomplete, Query Frontend calculates the required subqueries and executes them in parallel on downstream queriers. Query Frontend can optionally align queries with their step parameter to improve the cacheability of the query results. Currently, in-memory cache (fifo cache) and memcached are supported.

In-memory #

  max_size: ""
  max_size_items: 0
  validity: 0s

max_size: Maximum memory size of the cache in bytes. A unit suffix (KB, MB, GB) may be applied.

_NOTE: If both max_size and max_size_items are not set, then the cache would not be created.

If either of max_size or max_size_items is set, then there is not limit on other field. For example - only set max_size_item to 1000, then max_size is unlimited. Similarly, if only max_size is set, then max_size_items is unlimited.

Example configuration: kube-thanos

Memcached #

  addresses: []
  timeout: 0s
  max_idle_connections: 0
  max_async_concurrency: 0
  max_async_buffer_size: 0
  max_get_multi_concurrency: 0
  max_item_size: 0
  max_get_multi_batch_size: 0
  dns_provider_update_interval: 0s
  expiration: 0s

Slow Query Log #

Query Frontend supports --query-frontend.log-queries-longer-than flag to log queries running longer than some duration.

Naming #

Naming is hard :) Please check here to see why we chose query-frontend as the name.

Flags #

usage: thanos query-frontend [<flags>]

query frontend

  -h, --help                  Show context-sensitive help (also try --help-long
                              and --help-man).
      --version               Show application version.
      --log.level=info        Log filtering level.
      --log.format=logfmt     Log format to use. Possible options: logfmt or
                              Path to YAML file with tracing configuration. See
                              format details:
                              Alternative to 'tracing.config-file' flag (lower
                              priority). Content of YAML file with tracing
                              configuration. See format details:
                              Listen host:port for HTTP endpoints.
      --http-grace-period=2m  Time to wait after an interrupt received for HTTP
                              Split queries by an interval and execute in
                              parallel, it should be greater than 0 when
                              response-cache-config is configured.
                              Maximum number of retries for a single request;
                              beyond this, the downstream error is returned.
                              Limit the query time range (end - start time) in
                              the query-frontend, 0 disables it.
                              Maximum number of queries will be scheduled in
                              parallel by the Frontend.
                              Most recent allowed cacheable result, to prevent
                              caching very recent results that might still be in
                              Enable partial response for queries if no
                              partial_response param is specified.
                              --no-query-range.partial-response for disabling.
                              Path to YAML file that contains response cache
                              Alternative to
                              'query-range.response-cache-config-file' flag
                              (lower priority). Content of YAML file that
                              contains response cache configuration.
                              Use compression in results cache. Supported values
                              are: 'snappy' and ” (disable compression).
                              URL of downstream Prometheus Query compatible API.
                              Compress HTTP responses.
                              Log queries that are slower than the specified
                              duration. Set to 0 to disable. Set to < 0 to
                              enable on all queries.
                              Request Logging for logging the start and end of
                              requests. LogFinishCall is enabled by default.
                              LogFinishCall : Logs the finish call of the
                              requests. LogStartAndFinishCall : Logs the start
                              and finish call of the requests. NoLogCall :
                              Disable request logging.

Found a typo, inconsistency or missing information in our docs? Help us to improve Thanos documentation by proposing a fix on GitHub here ❤️