Sean Hanson comments

Results 38 comments of


                                            Sean Hanson

index lastUpdated timestamps are only stored on leaf nodes

I just meant doing a Depth-first search to try to find a leaf within the time boundaries could be very expensive for a large enough index.

request: tag-native divide and asPercent functions

Is this still being worked on?

request: tag-native divide and asPercent functions

I believe @robert-milan is doing work on this.

groupByTags tag parsing adds bad tags

Perhaps `groupByTags` should be "smarter" in building the key to prevent this issue instead of relying on string encoding/parsing?

groupByTags tag parsing adds bad tags

> @shanson7 do you feel the following 2 tests adequately cover the problem Maybe? I'd like to see those tests using the `name` tag and a case using another tag...

msgp: too few bytes left to read object

I'm seeing these fairly frequently. Any idea what the issue is?

msgp: too few bytes left to read object

I deployed a “silent node” (carbon in, partition 9999) and added some debug statements It turns out the buffers are coming back as nil `2018/05/09 21:05:26 [dataprocessor.go:223 func1()] [E] DEBUG...

msgp: too few bytes left to read object

I think this is ccache corruption. For this particular repro request it was always the same instance that was breaking things. I sent a ccache/delete request and now the error...

msgp: too few bytes left to read object

Sorry, to clarify: 1. The main issue of `msgp: too few bytes left to read object` is coming from [here](https://github.com/grafana/metrictank/blob/a19124e/api/dataprocessor.go#L221). This happens when the request to the peer is canceled...

Use datapoint timestamp to determine persist to cassandra idx

We had another customer complaining about this. They publish forecasted data every Sunday for the next week. Because of the compressed publishing window, the `lastUpdate` is roughly 7 days behind....