pipelinedb
pipelinedb copied to clipboard
High-performance time-series aggregation for PostgreSQL
When performing a schema only dump we receive the following warning: ``` pg_dump: [sorter] WARNING: could not resolve dependency loop among these items: pg_dump: [sorter] TABLE conversion_daily (ID 357 OID...
Having trouble using keyed_max with in view that joins a table with a stream in a subselect ```sql CREATE VIEW online_users WITH (action=materialize, ttl = '5 MINUTE', ttl_column='pinged_at') AS SELECT...
We [bail early](https://github.com/pipelinedb/pipelinedb/blob/master/src/combiner.c#L1205) during SW state initialization if the underlying query doesn't aggregate. This results in unexpected output stream write semantics.
With built in partitioning syntax [coming in PostgreSQL 10](https://www.depesz.com/2017/02/06/waiting-for-postgresql-10-implement-table-partitioning/), it would be great to be able to partition `CONTINUOUS VIEW`s. Currently you could partition with a `CONTINUOUS TRANSFORM` and a...
For partitioned CVs, the `reaper` should simply `DROP` partitions containing only TTL-expired rows. We should probably make this configurable, e.g.: ```sql CREATE VIEW v0 WITH (ttl = '1 day', ......
Would be great to have pipelinedb extensions available on PGXN and versioned!
Looking over the configuration settings for pipelinedb, I'm trying to figure out what is practical for some of these. I understand it is largely based on usage + throughput -...
i want to save the aggregate result from a view to one postgres table for example CREATE CONTINUOUS VIEW v1 as select sum(amount) as sum_amount,count(*) as all_count from some_steam and...
I'm working in a program that drops all views, drops all streams, and recreates them again on every run, so i executed this program hundreds of times, every time i...
```sql --DDL CREATE VIEW bi_monitor.rt_view_request_qps_stat WITH (action=materialize) AS SELECT log#>>'{c-host}' as request_domain, COUNT(*) as req_min, SUM(COALESCE(cast(log#>>'{c-bytes}' as numeric),0)) as rb_min, SUM(COALESCE(cast(log#>>'{s-bytes}' as numeric),0)) as sb_min, COUNT(DISTINCT (COALESCE(log#>>'{c-ip}',''),COALESCE(log#>>'{c-port}'),'')) as client_conns_min FROM...