Re: Which partition scheme makes sense for my time based IoT-datagrams?

On 2021-02-10 16:09:30 +0100, Thorsten Schöning wrote:
> I have a table storing datagrams from some IoT-devices, with one
> datagram per device per day most likely for around 75'000 devices
> currently. I want to test query performance with a partitioned table
> and am interested in the following queries mostly:
>
> * querying arbitrary datagrams by their ID
> * querying datagrams being X old based on some timestamp
> * 15 minutes

If you get one datagram per day, how it is useful to query all datagrams
received in the last 15 minutes? That's either a random sample (if
devices report at random times) or empty (if they all report at midnight
and it isn't just after midnight).

hp

--
_ | Peter J. Holzer | Story must make more sense than reality.
|_|_) | |
| | | hjp(at)hjp(dot)at | -- Charles Stross, "Creative writing
__/ | http://www.hjp.at/ | challenge!"