Use existing relational tables

[ftgo]

I am facing multiple performance issues with large datasets and high throughput. Basically it occurs because either the index is not present (default) and it takes too long to retrieve or because I define indexes (tried all kinds) and it takes too much time to recalculate.

I think AGE would leverage the combination with existing and optimized tables, such as using TimescaleDB hypertables as vertex.

I would like to understand what would stop AGE from following TimescaleDB approach, where you can do something like:

CREATE TABLE "conditions"(
    time         TIMESTAMP WITH TIME ZONE NOT NULL,
    device_id    TEXT,
    temperature  NUMERIC,
    humidity     NUMERIC
);

CREATE INDEX ON "conditions"(time DESC);
CREATE INDEX ON "conditions"(device_id, time DESC);
-- 86400000000 is in usecs and is equal to 1 day
SELECT create_hypertable('conditions', 'time', chunk_time_interval => 86400000000);

As you can see, the table creation process is detatched from the step to transform it to an hypertable. As long as you have a time column it is fine.

For AGE we could do the same, but instead of a time column we should require a agtype.