Unable to query more than 30 days using loki datasource

Grafana Loki
,
1

Hi,

I have tried everything I could on the loki side and can confirm than when I use the api I can return data that are over 1 year old.
However, when I use the time picker in Grafana and select last 90 days, I get the following error “the query time range exceeds the limit (query length: 2160h0m0s, limit: 30d1h)”.

Example of direct access with the api:

From 2024-01-01 00:00:00 UTC → 1704153600000000000
To 2025-01-01 00:00:00 UTC → 1735689600000000000

# curl -G \
  --data-urlencode 'query={job=~".+"}' \
  --data-urlencode 'limit=100' \
  --data-urlencode 'start=1704153600000000000' \
  --data-urlencode 'end=1735689600000000000' \
  'http://localhost:3100/loki/api/v1/query_range'

I get the following result (truncated for privacy):

{"status":"success","data":{"resultType":"streams","result":[{"stream":{"client":"client1","detected_level":"info","filename":"/data/ingester/client/server1/web.log","job":"client-logs","...

In the logs I can clearly see the request:

Jul 22 10:54:47 Europa loki[14327]: level=info ts=2025-07-22T10:54:47.059538937Z caller=roundtrip.go:379 org_id=fake traceID=075c8e35899931fd msg="executing query" type=range query="{job=~\".+\"}" start=2024-01-02T00:00:00Z end=2025-01-01T00:00:00Z start_delta=13618h54m47.059534186s end_delta=4858h54m47.059534371s length=8760h0m0s step=126144000 query_hash=453119268
Jul 22 10:54:47 Europa loki[14327]: level=debug ts=2025-07-22T10:54:47.061741905Z caller=shard_resolver.go:123 org_id=fake traceID=075c8e35899931fd bytes=393kB chunks=18 streams=7 entries=1792 msg="queried index" type=single matchers="{job=~\".+\"}" duration=99.08µs from=2024-12-31T22:59:30Z through=2025-01-01T00:00:00Z length=1h0m30s

However, when I use Grafana, it looks like the request does not even reach loki. In case the above error would be coming from loki, I believe we should see something in debug logs right?

Here is my loki config :

auth_enabled: false

server:
  http_listen_port: 3100
  grpc_listen_port: 9096
  log_level: debug
  grpc_server_max_concurrent_streams: 1000

common:
  instance_addr: 127.0.0.1
  path_prefix: /tmp/loki
  replication_factor: 1
  ring:
    kvstore:
      store: memberlist

memberlist:
  join_members: []

table_manager:
  retention_deletes_enabled: true
  retention_period: 8760h

storage_config:
  azure:
    account_key: ==
    account_name: loki
    container_name: chunks
    use_managed_identity: false
    request_timeout: 0

query_range:
  results_cache:
    cache:
      embedded_cache:
        enabled: true
        max_size_mb: 100

limits_config:
  metric_aggregation_enabled: true
  enable_multi_variant_queries: true
  max_query_length: 0h
  max_query_lookback: 87600h
  volume_enabled: true

schema_config:
  configs:
    - from: 2020-10-24
      store: tsdb
      object_store: azure
      schema: v13
      index:
        prefix: index_
        period: 24h

ingester:
  max_chunk_age: 4320h

pattern_ingester:
  enabled: true
  metric_aggregation:
    loki_address: localhost:3100

ruler:
  alertmanager_url: http://localhost:9093

frontend:
  encoding: protobuf

compactor:
  compaction_interval: 2h # or 1h
  working_directory: /data/loki/retention
  retention_enabled: false
  delete_request_store: "azure"
  retention_delete_delay: 5m
  retention_delete_worker_count: 150
  delete_request_cancel_period: 5m

I believe that loki works as expected but please let me know if you see something missing/wrong in the config.
Also, is there something specific in grafana to be set to allow more than 30 days?

Versions:

# /usr/share/grafana/bin/grafana -version
grafana version 12.0.2+security-01

# /usr/bin/loki --version
loki, version 3.5.2 (branch: release-3.5.x, revision: 257d2f62)
  build user:       root@67bdab5ad0d6
  build date:       2025-07-10T19:26:46Z
  go version:       go1.24.1
  platform:         linux/amd64
  tags:             netgo

I hope someone can help.
Thanks a lot in advance.

2

I think it’s probably the duration of the query, not how far back it is.

Your curl command shows a difference of 24 hours between start and end time, so your query duration is 24h. But if you pick 90 days in Grafana you would be sending a query that’s from 90 days in the past to now, which is longer than your configured duration of 30d.

1 Like
3

Thanks a lot for your reply.

Not sure I understand, as far as i can tell my query in curl should be for the following dates.

1. 1704153600000000000 nanoseconds

Convert to seconds:
1704153600000000000 / 1_000_000_000 = 1704153600 seconds

Then convert to a human-readable date:
:spiral_calendar: January 2, 2024, at 00:00:00 UTC

2. 1735689600000000000 nanoseconds

Convert to seconds:
1735689600000000000 / 1_000_000_000 = 1735689600 seconds

Human-readable date:
:spiral_calendar: November 1, 2024, at 00:00:00 UTC

Why do you say I am asking only for 24h ?

=> I see. You checked the logs. those might not be the correct. Let me try again and share the correct logs.