Hi
We have a scenario that we can not really figure out how to solve. We log a line 13 times basically the same time during the startup of an application, “Doing a thing”.
This line get logged so fast that that all 13 lines gets somewhere between 4 and 5 unique timestamps since they are being run at the same time and we can se that there is 13 lines being sent to Loki if we trace the traffic using a proxy.
Loki says nothing about these lines in its logs even with debug logs but when we query them we only get 4-5 results returned, we did turn on increment_duplicate_timestamp since we suspected that it might be the reason but we still only get 4-5 lines each time even if the application always log 13.
We did try to add a guid to each line just to get a unique line and then we see 13 lines in Loki as expected.
We have tested against a local docker instance without any additional configuration as well as our production Loki cluster in microservice mode with the same result.
Does anyone have an idea what could cause this and how to fix it? The log lines are important events we need to keep track of so we need to see all 13.