compact: Add compactor time validation

CHANGELOG.md

@@ -25,6 +25,7 @@ We use *breaking :warning:* to mark changes that are not backward compatible (re
 
 ### Added
 
+- [#7293](https://github.com/thanos-io/thanos/pull/7293) Compact: Add validation to --min-time and --max-time flags
 - [#7231](https://github.com/thanos-io/thanos/pull/7231) Tracing: added missing sampler types
 - [#7194](https://github.com/thanos-io/thanos/pull/7194) Downsample: retry objstore related errors
 - [#7105](https://github.com/thanos-io/thanos/pull/7105) Rule: add flag `--query.enable-x-functions` to allow usage of extended promql functions (xrate, xincrease, xdelta) in loaded rules

cmd/thanos/compact.go

@@ -228,6 +228,10 @@ func runCompact(
 		}
 	}()
 
+	if err := validateFilterConf(conf.filterConf); err != nil {
+		return fmt.Errorf("invalid time filters: %s", err.Error())
+	}
+
 	// While fetching blocks, we filter out blocks that were marked for deletion by using IgnoreDeletionMarkFilter.
 	// The delay of deleteDelay/2 is added to ensure we fetch blocks that are meant to be deleted but do not have a replacement yet.
 	// This is to make sure compactor will not accidentally perform compactions with gap instead.
@@ -689,6 +693,27 @@ func runCompact(
 	return nil
 }
 
+func validateFilterConf(filterConf *store.FilterConfig) error {
+	if filterConf.MinTime.PrometheusTimestamp() > filterConf.MaxTime.PrometheusTimestamp() {
+		return errors.New("max time must be a time or duration after min time")
+	}
+
+	if err := checkIfDurInFuture(filterConf.MinTime.Dur); err != nil {
+		return fmt.Errorf("min time: %s", err.Error())
+	}
+	return nil
+}
+
+func checkIfDurInFuture(dur *model.Duration) error {
+	if dur == nil {
+		return nil
+	}
+	if !strings.HasPrefix("-", dur.String()) {
+		return errors.New("duration cannot be in the future")
+	}
+	return nil
+}
+
 type compactConfig struct {
 	haltOnError                                    bool
 	acceptMalformedIndex                           bool
@@ -823,9 +848,9 @@ func (cc *compactConfig) registerFlag(cmd extkingpin.FlagClause) {
 		Default("").EnumVar(&cc.hashFunc, "SHA256", "")
 
 	cc.filterConf = &store.FilterConfig{}
-	cmd.Flag("min-time", "Start of time range limit to compact. Thanos Compactor will compact only blocks, which happened later than this value. Option can be a constant time in RFC3339 format or time duration relative to current time, such as -1d or 2h45m. Valid duration units are ms, s, m, h, d, w, y.").
+	cmd.Flag("min-time", "Start of time range limit to compact. Thanos Compactor will only compact blocks whose max time is at a date more recent than this value. Option can be a constant time in RFC3339 format or time duration relative to current time, such as -1d or -2h45m. Durations must be negative, and valid units are ms, s, m, h, d, w, y.").
 		Default("0000-01-01T00:00:00Z").SetValue(&cc.filterConf.MinTime)
-	cmd.Flag("max-time", "End of time range limit to compact. Thanos Compactor will compact only blocks, which happened earlier than this value. Option can be a constant time in RFC3339 format or time duration relative to current time, such as -1d or 2h45m. Valid duration units are ms, s, m, h, d, w, y.").
+	cmd.Flag("max-time", "End of time range limit to compact. Thanos Compactor will only compact blocks whose min time is at a date less recent this value. Option can be a constant time in RFC3339 format or time duration relative to current time, such as -1d or -2h45m. Valid units for duration are ms, s, m, h, d, w, y.").
 		Default("9999-12-31T23:59:59Z").SetValue(&cc.filterConf.MaxTime)
 
 	cmd.Flag("web.disable", "Disable Block Viewer UI.").Default("false").BoolVar(&cc.disableWeb)

docs/components/compact.md

@@ -251,7 +251,13 @@ Generally there two scalability directions:
 
 1. Too many producers/sources (e.g Prometheus-es) are uploading to same object storage. Too many "streams" of work for Compactor. Compactor has to scale with the number of producers in the bucket.
 
-You should horizontally scale Compactor to cope with this using [label sharding](../sharding.md#compactor). This allows to assign multiple streams to each instance of compactor.
+You can horizontally scale Compactor to cope with this via
+- [label sharding](../sharding.md#compactor): This allows to assign multiple streams to each instance of compactor.
+- [time sharding](../sharding.md#compactor): This allows for compactors to only handle blocks that fall within a given time range
+
+For example, to have an instance of compactor only working on blocks that fall within the time range of 14 days ago to 7 days ago, the flags would be `--min-time=-14d --max-time=-7d`
+
+Absolute times can be specified as well.
 
 2. TSDB blocks from single stream is too big, it takes too much time or resources.
 
@@ -399,20 +405,21 @@ Flags:
                                 json.
       --log.level=info          Log filtering level.
       --max-time=9999-12-31T23:59:59Z
-                                End of time range limit to compact.
-                                Thanos Compactor will compact only blocks,
-                                which happened earlier than this value. Option
+                                End of time range limit to compact. Thanos
+                                Compactor will only compact blocks whose min
+                                time is at a date less recent this value. Option
                                 can be a constant time in RFC3339 format or time
                                 duration relative to current time, such as -1d
-                                or 2h45m. Valid duration units are ms, s, m, h,
-                                d, w, y.
+                                or -2h45m. Valid units for duration are ms, s,
+                                m, h, d, w, y.
       --min-time=0000-01-01T00:00:00Z
-                                Start of time range limit to compact.
-                                Thanos Compactor will compact only blocks, which
-                                happened later than this value. Option can be a
-                                constant time in RFC3339 format or time duration
-                                relative to current time, such as -1d or 2h45m.
-                                Valid duration units are ms, s, m, h, d, w, y.
+                                Start of time range limit to compact. Thanos
+                                Compactor will only compact blocks whose max
+                                time is at a date more recent than this value.
+                                Option can be a constant time in RFC3339 format
+                                or time duration relative to current time, such
+                                as -1d or -2h45m. Durations must be negative,
+                                and valid units are ms, s, m, h, d, w, y.
       --objstore.config=<content>
                                 Alternative to 'objstore.config-file'
                                 flag (mutually exclusive). Content of