3 ms·

Curious to know if any TSDBs (Timeseries databases) evaluated before finalizing on Cassandra to store the traces.

by mnsmns 6y ago

user5994461 6y ago
I don't think there's any free TSDB available. The project is maybe 3 years old. InfluxDB is more recent and quite limited. Things like sharding is not supported and they stated it would never be supported except in a paid edition when they make one. Prometheus is more recent. Similar story with scaling. They changed storage formats and rewrote once or twice in the past few years, it's moving really fast. It's more of a standalone product for server metrics (node exporter + prometheus + grafana), wouldn't recommend to use as a general purpose database.
- 02020202 6y ago
  s you're saying that big tech does not want to PAY for using big tech? :D
- mnsmns 6y ago
  As of today sharding is supported in InfluxDB. Prometheus is not suitable for traces, it is more for metrics. The problem I see with all these DBs are they are well suited for specific applications like prometheus for metrics, Influx for IoT. Timescale is still in nascent stages.
Thaxll 6y ago
Also Netflix is a Java shop so anything JVM based makes sense for them.
- geodel 6y ago
  I feel in between Linkedin, Twitter and Netflix all Java design pattern/framework/architect experts are located. A pleasant surprise to me is Oracle seems to have gotten rid of such products and people.
jrockway 6y ago
I don't think time is a particularly strong keying methodology for distributed tracing. You want to be able to quickly select all spans that have a certain trace ID. The time that the span was ingested is relatively unimportant to the querying mechanics, and you don't do any operations like "combine samples into a 1h average after 5d".
- jeffbee 6y ago
  When I was at Dropbox I did do this ... we only retained traces for a very short time, so I did daily aggregations of traces into summary data structures that could attribute service time to various span types at different depths of the request etc. Don't ignore the value of aggregating and summarizing trace data. Of course, a TSDB is also useless for this, I agree there.
  - jrockway 6y ago
    Yeah, using the trace data to get RPC latency and success/error charts is useful, but that's just another branch off the data-stream. Your metrics system needs to inspect every trace, so your database only needs the ability to send you a list. (Jaeger seems to have Kafka support so you can do this in real time. Haven't tried it.)
ecnahc515 6y ago
For what it's worth, Jaeger, one of the most popular FOSS tracing servers uses ES/Cassandra for storing it's traces.

Curious to know if any TSDBs (Timeseries databases) evaluated before finalizing on Cassandra to store the traces.

by mnsmns 6y ago

Curious to know if any TSDBs (Timeseries databases) evaluated before finalizing on Cassandra to store the traces.

I don't think there's any free TSDB available. The project is maybe 3 years old. InfluxDB is more recent and quite limited. Things like sharding is not supported and they stated it would never be supported except in a paid edition when they make one. Prometheus is more recent. Similar story with scaling. They changed storage formats and rewrote once or twice in the past few years, it's moving really fast. It's more of a standalone product for server metrics (node exporter + prometheus + grafana), wouldn't recommend to use as a general purpose database.

02020202 6y ago
s you're saying that big tech does not want to PAY for using big tech? :D
mnsmns 6y ago
As of today sharding is supported in InfluxDB. Prometheus is not suitable for traces, it is more for metrics. The problem I see with all these DBs are they are well suited for specific applications like prometheus for metrics, Influx for IoT. Timescale is still in nascent stages.

Thaxll 6y ago

Also Netflix is a Java shop so anything JVM based makes sense for them.

geodel 6y ago
I feel in between Linkedin, Twitter and Netflix all Java design pattern/framework/architect experts are located. A pleasant surprise to me is Oracle seems to have gotten rid of such products and people.

jrockway 6y ago

I don't think time is a particularly strong keying methodology for distributed tracing. You want to be able to quickly select all spans that have a certain trace ID. The time that the span was ingested is relatively unimportant to the querying mechanics, and you don't do any operations like "combine samples into a 1h average after 5d".

jeffbee 6y ago
When I was at Dropbox I did do this ... we only retained traces for a very short time, so I did daily aggregations of traces into summary data structures that could attribute service time to various span types at different depths of the request etc. Don't ignore the value of aggregating and summarizing trace data. Of course, a TSDB is also useless for this, I agree there.
- jrockway 6y ago
  Yeah, using the trace data to get RPC latency and success/error charts is useful, but that's just another branch off the data-stream. Your metrics system needs to inspect every trace, so your database only needs the ability to send you a list. (Jaeger seems to have Kafka support so you can do this in real time. Haven't tried it.)

ecnahc515 6y ago

For what it's worth, Jaeger, one of the most popular FOSS tracing servers uses ES/Cassandra for storing it's traces.