Partitioning questions
I have a couple of tables that are pretty big, 2+ billion rows. Now, querying by time is fast and there is no problem. But sometimes I want to use these tables to make queries where I need to, for...
View ArticleSkipScan node with Distinct on not working when using an `in` clause
I am attempting to use a DISTINCT ON query, but want to use the in clause. I see the SkipScan piece in the explain analyze ONLY when using an =, but fails to use SkipScan when using an in. Here’s the...
View ArticleClik here to view.
External trigger after each continuous aggregate finished?
I watched the new “Timecale for Finance” webinar episodes on youtube. In connection with the hierarchical continuous aggregates, I had the following question: Let’s assume I continuously write trades...
View ArticleClik here to view.
Slow select from compressed hypertable by identifier from PK
Hi everybody! Once upon a time I compress hypertables. And it was the end… of the lightning fast select of 2+ id values which corresponds to PK. The end of story ) Question: how should I write queries...
View ArticleHow to move data from one hypertable to new design hypertable?
Using TimescaleDB v2.14.2 self-hosted on prem on PostgreSQL 15.6 on Red Hat 9.3. Software is running on vmware virtual machine with 16 CPU-cores and 256 GB of RAM and few 10s of TB of disk. We have...
View ArticleUpdate Performance
Hello, we are facing troubles with insert and update of values in our Timescale table. In our db we have databases for each machine (~20max) and each machine has a table which contains minutely data...
View ArticleOptimizing telegraf with timescaledb
I use telegraf with timescaledb and it works great. I use this schema – telegraf/plugins/outputs/postgresql at v1.30.1 · influxdata/telegraf · GitHub I was wondering if there are any optimizations...
View ArticleEstimated release date for 2.15
Do you have an estimate in terms of release date for 2.15? We have a special interest updates related to indexing of compressed chunks. 1 post - 1 participant Read full topic
View ArticleTimescaledb on tmpfs
This is just poking into relatively undiscovered territory from my side, but would it be possible to persist timescale data on tmpfs filesystem, essentially making timescale an in-memory database? 1...
View ArticleRecalculate materialized table after dropping chunks
I have a hypertable: measurements, which is materialized into: measurements_hourly. I accidentally dropped chunks from meaurement_hourly instead measurements by running: SELECT...
View ArticlePostgresql pg_upgradecluster error
Hi ! I’m trying to upgrade my postgresql 15 cluster to postgresql 16 with the pg_upgradecluster command. Knowing that I have a main node and 2 replicas with synchronous streaming. When I tried to run...
View ArticleHow to revert back to normal table from hypertable
Hello, I created hypertable and saw a performance decline because many chunks were created (1 sec interval). This was done automatically because I did not specify INTERVAL. The documents however says...
View ArticleLast(*) in timebucket
Hey, let me lay out my scenario. I have a bunch of data, each row is around 1-2 seconds apart and there is approx 10 million rows of data. Each rows has an account_id and value. For each account I...
View ArticleTimeScale only uses a single core in compression?
I have tested compressing chunks manually in TimeScale on premise using: select compress_chunk(i, if_not_compressed => true) from show_chunks(‘table’) i; Question: The above will compress all...
View ArticleDuration_in over multiple values
I have a column of data that can take on about 20 enumerated values. I am looking for the total time the parameter is in 5 of these states. I have already started using duration_in(state_agg(dtm,...
View ArticleHyperfunction for time meter
Is there a hyperfunction that somehow can be used to retrieve state time? In this case, I have a events table for machine states. It is supposed to be a insert only table, so when machine A changes to...
View ArticleConnections pool
Hello My team has a large timescale database, from which we query often, and some of the queries might take up to 5-7 seconds. We use the python databases package to connect to the db, with...
View Article`last` function when the timestamp is the same
Hey all, Say i have a dataset of 100,000,000 rows, many of which share the same timestamp. If i call the last function on them, say last(id, timestamp), is there deterministic behaviour to determine...
View ArticleTimescaledb 2.14.2 crashed for queries that have a large time span
We are using timescaledb 2.14.2 with postgres 15.6 (Ubuntu 15.6-1.pgdg22.04+1). we have a table that have data retention of 24 hours. When we query the table with a time span that is larger than 2...
View ArticleHypertable_size() does not return correct number
Hello, We use TimescaleDB as our storage in our project. We have a hypertable organized as daily chunks and we have a functionality to dump data from this table to csv. This dump functionality works...
View Article