-
Notifications
You must be signed in to change notification settings - Fork 3.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Time range queries do not work when boltdb-shipper is used #2816
Comments
I investigated the issue a little further and I can provide additional information. The
|
Thanks @KaloyanSimitchiyski this is very helpful info! we will see what we can figure out. |
Just wanted to say I'm seeing the same thing. I'm using Azure blob storage, same behavior. Thanks! |
@slim-bean Will take a look! |
I have opened a PR to fix the issue. I feel there should be a problem only when query start time is withing last ~1.5h(for both relative and absolute time ranges). If the start time is before that then there should not be a problem. |
Hi Folks, Is this fix on the Release 2.0 which is available for download? As far as I can see on the Release page, it dates from Oct 26, but this fix seems to come few days later. Is there any place where I could download it already compiled or on that case I should compile it on my end? BR |
Bug Description
Hello! Since upgrading from 1.6.1 to 2.0.0, any time range queries that I perform from Grafana (or directly from the Loki API) return all logs up until the configured max lines for the data source (e.g. 1000). The specified time range is not taken into account at all. The problem occurs for both relative and absolute time ranges. I'm running a single instance of monolith Loki 2.0.0 with GCS as a chunk store and boltdb-shipper as an index store. This setup worked fine in 1.6.1. If instead of boltdb-shipper, I switch to the vanilla boltdb, everything works as expected even on 2.0.0. During my testing, I deployed a clean Loki (empty bucket and persistent volumes) every time I changed the chunk/index store types.
To Reproduce
Steps to reproduce the behavior:
Expected behavior
Time ranged queries to work properly. Only logs that fall between the specified times should be displayed.
Environment:
Promtail config is the same as in the official helm chart. This is my Loki config:
I executed a range query while using boltdb-shipper and while using vanilla boltdb. Here are logs from Loki and request/response from the Grafana query inspector:
Logs when using GCS + boltdb-shipper (problem occurs)
Loki logs:
Query Inspector (Notice how only the first log message falls in the specified range):
Logs when using vanilla boltdb (problem does not occur)
Loki logs:
Query Inspector (all messages are in the specified range):
Please, let me know if you need more information.
Best regards,
Kaloyan
The text was updated successfully, but these errors were encountered: