Riley Zimmerman

Optionally disable metric_counters

3

In the initial CQL release, the metric_counters were broken: https://github.com/kairosdb/kairosdb/issues/451 Now that they're back in 1.2.1 they're again adding a small but noticeable overhead on my environment. The exact amount...

Change batch_size to a limit based on bytes to match Cassandra fail threshold

1

KairosDB uses `kairosdb.queue_processor.batch_size`, which is the max number of messages it will send in a batch. These can have varying sizes in bytes. Cassandra uses `batch_size_fail_threshold_in_kb`. I set both to...

Feature request: Handle multiple datapoints tables to utilize TWCS with multiple TTLs

2

Time Window Compaction Strategy (TWCS) offers large advantages over Size Tier Compaction Strategy when working with time series data. The data is bucketed into windows as it comes in. Once...

Datastore writing errors after moving from KairosDB 1.1.0 to 1.1.3

15

I am using Cassandra 3.9. With KairosDB 1.1.0 I have been ingesting ~1 mill metrics per minute sustained, 3 million peak (after restarts when data has queued up). I have...

SSTable max local deletion time allowing for missed deletion?

5

Hi, I've been testing with TWCS and KairosDB. My KairosDB TTL for data is 15 days. Here is the SCHEMA (note the 'timestamp_resolution': 'MILLISECONDS'): ``` CREATE TABLE metricdb.data_points ( key...