Scalable Time Series With Cassandra

Cassandra is awesome for many things. One of the things it's awesome for is Time Series. Combining the power of Cassandra with APIs of existing Time Series tools, such as Graphite can yield interesting results.

Cyanite is a Time Series aggregator and store built on top of Cassandra. It's fully compatible with Graphite, can serve as a plug-in replacement for Graphite and Graphite web.

Cyanite is using SASI indexes to make glob metric path queries, can query and aggregate, store, display and analyse metrics from hundreds and thousands of servers.

Which data modelling practices work best for Time Series, which new awesome Cassandra features you can use to make your Time Series analysis better.

Transcript

1. Alex Petrov Scalable time-series applications with Cassandra
2. Cyanite past, present and future
3. Requirements
4. Throughput Incoming data Aggregates Read queries
5. Scalability Paths / metrics Historical data Readers / writers
6. vs raw Aggregated
7. Predefined list of reports to aggregate Reducing amount of data points Faster queries Precision loss due to aggregation
8. “servers.*.workers.busyWorkers”: “10s:7d,1m:21d,15m:5y”
9. Graphite
10. Store numeric time-series data Render graphs of this data on demand https://graphiteapp.org/
11. Ecosystem
12. carbon: listens for time-series data whisper: DB for storing TS data graphite-web: UI & API for graphs
15. Downsides Single-host solution Plenty of disk seeks Optimised for space Sharding / replication is “manual”
16. Scaling Cluster topology stored on every node Manual replication Changing cluster topology non-trivial
17. Scaling Stateless service Automatic shard assignment Replication Easy management Easy topology changes
18. reminds you of anything?
19. Cassandra perfect match
20. Cyanite Stateless Async I/O Custom scheduler No Whisper files Distributed Horizontally scalable
21. Cyanite responsibilities Carbon-compatible listener Aggregate data in-memory Flush aggregates to Cassandra Path storage Retrieve paths for query Aggregate the query results
22. Cassandra features Metric expiry, TTL User Defined Types SASI Indexes, LIKE queries Aggregate Functions Offline data loading Clustering range queries IN partition key locking
23. Globbing index
24. app.cluster.server.subsystem.metric
25. Built with SASI indexes Fast, scalable queries Optimised for glob
26. CREATE TABLE segment ( parent text, segment text, pos int, length int, leaf boolean, PRIMARY KEY (parent, segment))
27. app.cluster.server.subsystem.metric
28. SELECT * FROM segments WHERE parent = 'root' AND pos = 1 Wildcard: *
29. SELECT * FROM segments WHERE pos = 4 Postfix: *.*.*.*.metric
30. SELECT * FROM segments WHERE parent = 'app' AND pos = 2 ALLOW FILTERING Prefix: app.*
31. SELECT * FROM SEGMENTS WHERE pos = 3 AND segment LIKE 'abc.%' ALLOW FILTERING Suffix: abc.*.metric
32. engine Query
33. 2-Step transformations: Inner Outer Cross metric engine Query
34. a.b.c1 a.b.c2 scale(a.b.*, 10.0) {"a.b.c1" [1 2 3] "a.b.c2" [5 6 7]} {"a.b.c1" [10 20 30] "a.b.c2" [50 60 70]}
35. a.b.c derivative(a.b.c) {"a.b.c" [1 3 6]} {“derivative(a.b.c)" [nil 2 3]}
36. a.b.c1 a.b.c2 sumSeries(a.b.c1,a.b.c2) {"a.b.c1" [1 1 1] "a.b.c2" [2 2 2]} {“sumSeries(a.b.c,a.b.d)" [3 3 3]}
37. model Data
38. CREATE TYPE IF NOT EXISTS metric_resolution ( precision int, period int ); CREATE TYPE IF NOT EXISTS metric_id ( path text, resolution frozen<metric_resolution> );
39. CREATE TYPE IF NOT EXISTS metric_point ( max double, mean double, min double, sum double );
40. Scaling
41. Scaling cyanite Use DTSC Cyanite is stateless For high loads, split readers and writers Load-balance with HAProxy Colocate Cyanite & Cassandra
42. soon Coming
43. P-Square histograms T-Digest quantiles Gorilla Compression Cassandra aggregates Custom glob indexes Scheduler soon Coming
44. Kafka ingester Statsd protocol support Standalone cyanite Dynamic thresholds soon Coming