I have a very connection heavy multi-tenancy schema-based solution, and i would like more info on how connections scale with resources on TimeScale Cloud.
On TimeScale Cloud, how does the max_connections range scale with resources, on a single node instance?
I have tried a small 0.25 cpu which has a max_connections range of 25 to 25.
I have also tried one cpu with a range of 25 to 500 connections.
Hello @MartinMogh welcome to the forum and thank you for the question.
I’ve moved it to the Timescale Cloud category as the question’s specific to Timescale’s hosted services. First of all, can I check if you found this page in the docs? https://docs.timescale.com/cloud/latest/scaling-a-service/ I think probably what you need insight about is connection pooling… I found an article about MST connection pools but I am still looking for the equivalent cloud one… it may be useful background though?
From what you say it seems like you have a trial or paid account? The very best place to get the answer you need will be from our support team – you can access support from your Cloud console or by emailing [email protected] Also, if you didn’t have an intro session yet with a Customer Success Manager, let’s see if we can get that set so that you can talk with them about your use case and make sure you’re getting the right information – they can call in engineer(s) as necessary to talk with you. Please send me a direct message if you would like me to help set that up (I’m Timescale’s community manager).
All that said! If I can get a generic response here in the public forum, I will as I am sure it’s a question that will arise again
I am already using MST but am looking to change to timescale cloud.
What i am looking for is something like (but for TimeScale Cloud):
Connection Limits | Managed Service for TimescaleDB Knowledge Base
The MST connection pooling is not relevant, because we use client-side connection pool and not pgbouncer. Do not know if TimeScale Cloud supports pgbouncer.
Yes I want to know if pgbouncer is used in cloud. Currently using it in MST but I can’t see anything analogous in TSC.
I have a serverless application (Azure functions) which can scale to many connections in a short time and need the pooling on the server side
I think i asked the same question to the Timescale support guys, and the reply was:
In regards to migrating to Timescale Cloud we are glad you are interested on this. For connection pooling on Timescale cloud as of now it is not yet supported but it is part of the roadmap for Q2 2023 and should be available in the future