grafana / loki

Like Prometheus, but for logs.
https://grafana.com/loki
GNU Affero General Public License v3.0
23.82k stars 3.43k forks source link

Run the same query loki but different result output #8647

Open dtnguyen3011 opened 1 year ago

dtnguyen3011 commented 1 year ago

The fist run query -> Line limit: 5000 (46 returned) The second run query -> Line limit:5000 (18 returned)

This is my loki.yaml:

auth_enabled: false

server: http_listen_port: 3100 grpc_listen_port: 9096 http_server_read_timeout: 3m http_server_write_timeout: 3m

chunk_store_config: max_look_back_period: 0s

compactor: shared_store: filesystem working_directory: /data/loki/boltdb-shipper-compactor

memberlist: join_members:

distributor: ring: kvstore: store: memberlist

common: instance_addr: 127.0.0.1 path_prefix: /data/loki storage: filesystem: chunks_directory: /data/loki/chunks rules_directory: /data/loki/rules replication_factor: 10 ring: kvstore: store: memberlist

ingester: chunk_block_size: 1048576 chunk_idle_period: 0.5m chunk_retain_period: 1m

ruler: ring: kvstore: store: memberlist

query_range: results_cache: cache: embedded_cache: enabled: true max_size_mb: 50000 parallelise_shardable_queries: true

schema_config: configs:

storage_config: boltdb_shipper: active_index_directory: /data/loki/boltdb-shipper-active cache_location: /data/loki/boltdb-shipper-cache cache_ttl: 24h shared_store: filesystem

table_manager: retention_deletes_enabled: false retention_period: 0s

query_scheduler: max_outstanding_requests_per_tenant: 10000

frontend: max_outstanding_per_tenant: 10000

limits_config: enforce_metric_name: false max_entries_limit_per_query: 100000

Pls help me to check on this?

Many thanks

mlonV commented 1 year ago

hi,Is this a loki cluster? It's not a k8s

dtnguyen3011 commented 1 year ago

hi @mlonV my loki running in AKS with 3 pod loki container with loki grafana/loki:2.7.3 image

maxramqvist commented 1 year ago

I think we have the same issue. Loki 2.7.0 running on 2 VMs, with Minio as storage. When querying the Loki cluster we get different results depending on which instance the query ends up on. The same instance always returns the same result, though. I can't see anything seemingly related in the Loki logs.

brigadire commented 7 months ago

We also have a similar problem . We are using loki:2.6.1 version with s3 storage.