I am trying to do what should be a simple join but the tables are very large and it is taking a long, long time. I have the feeling that I have stuffed up something in the syntax.
Here is what I have:
telemetry=> select (tq1.timestamp = tq2.timestamp) as timestamp, tq1.value as q1, tq2.value as q2 from cal_quat_1 tq1 inner join cal_quat_2 as tq2 using (timestamp) where timestamp > '2004-01-12 09:47:56.0000 +0' and timestamp < '2004-01-12 09:50:44.7187 +0' order by timestamp;
telemetry=> \d cal_quat_1
Table "cal_quat_1"
Column | Type | Modifiers
-----------+--------------------------+-----------
timestamp | timestamp with time zone |
value | double precision |telemetry=> \d cal_quat_2
Table "cal_quat_2"
Column | Type | Modifiers
-----------+--------------------------+-----------
timestamp | timestamp with time zone |
value | double precision |My understanding of an inner join is that the query above will restrict this to finding tq1.timestamp, tq1.value and then move onto t12.value to search the subset. I have tried this with and without the '=' sign and it isn't clear if it is making any difference at all. I have not allowed the query to finish as it seems to take more than 10 minutes. Both timestamps are indexed and I expect about 150 rows to be returned. At the end of the day, I have four identical tables of quaternions (timestamp, value) and I need to extract them all for a range of timestamps.
Cheers, Randall
---------------------------(end of broadcast)--------------------------- TIP 8: explain analyze is your friend
