feat(ingester): Add owned series tracking to prevent false throttling…

[anant10]

What this PR does:

Adds per-ingester series ownership tracking to prevent false throttling during ingester scale-up and ring resharding.

The problem: When ingesters scale up, the per-ingester local series limit drops immediately (recalculated based on new ingester count), but stale series data remains in the TSDB head for up to 2 hours until head compaction. PreCreation() uses Head().NumSeries() for limit checks, so it incorrectly rejects new writes during this window — the ingester appears over its new lower limit, but many of those series have been resharded to other ingesters and will be cleaned up at next compaction.

The solution: Track which series each ingester actually owns according to the ring, and use that count for limit enforcement. The owned count drops immediately when the ring changes (within 1 minute), eliminating the 2-hour dependency on head compaction.

How it works:

1. On each push, compute the series' ring token (same hash the distributor uses for routing) and store it in ActiveSeries
2. Every ~1 min (updateActiveSeries cycle), if the ring changed, re-scan all entries and remove series whose token no longer maps to this ingester
3. PreCreation() uses activeSeries.Owned() instead of Head().NumSeries() for the limit check

Design decisions:

• Two feature flags for safe progressive rollout:
  • -ingester.owned-series-metrics-enabled: enables cortex_ingester_owned_series metric emission only (no enforcement change)
  • -ingester.owned-series-limit-enforcement-enabled: switches limit enforcement to use owned count (requires first flag)
• Zone-local ownership check (3 lines): SearchToken(zoneTokens, key) → is responsible token in this instance's set?
• Lock-free reads on push path: ring state stored behind atomic.Pointer[ringState] — zero lock contention
• Instance-level max_series: instanceOwnedCount recalculated every ~1 min (not incremental, avoids drift). Startup fallback to instanceSeriesCount when count is 0
• Code consolidation: FNV hash → pkg/util/fnv.go, sharding functions → pkg/ring/token.go (eliminates duplication between distributor and ring)

Validation:

Tested in a multi-zone deployment under sustained write load. After scale-up:

• memory_series on old ingesters remained high (stale data in head)
• owned_series on old ingesters dropped proportionally to ring redistribution
• Zero throttle errors during the scale-up window that would have previously caused false rejections

New configuration flags (experimental):

yaml

ingester:
 # Emit cortex_ingester_owned_series metric (no enforcement change)
 # CLI flag: -ingester.owned-series-metrics-enabled
 [owned_series_metrics_enabled: <bool> | default = false]

 # Use owned count for limit enforcement (requires above flag)
 # CLI flag: -ingester.owned-series-limit-enforcement-enabled
 [owned_series_limit_enforcement_enabled: <bool> | default = false]

Which issue(s) this PR fixes: Fixes #7509