I have two PostgreSQL database servers, one uses SSD, the other HDD.
We struggled with the full SSD disc and we didn't need all the data on SSD all the time so we moved the majority of them to the HDD.
Sometimes we need to work with some of the data again on the SSD to get some analysis done quickly (in hours instead of days). We think the most elegant solution would be just to alter the schema of a table of our choice on the HDD PostgreSQL to the replication schema, which would automatically move this table to the SSD PostgreSQL. We then would do our work in another SSD schema and, by moving the results again to the SSD PostgreSQL synchronized schema, everything would sync back to HDD.
We would like to start replication of that schema either after some interval (e.g. at midnight) or by calling some function when we have finished our work. Is there an easy way to solve this problem? Is the described approach correct in the first place? We don't want to implement custom scripts.