Memory pressure
Needs attention86%
Share of reads answered from memory instead of disk.
Healthy databases sit above 99%. Below that, everyday queries wait on disk because the data no longer fits in memory.
How much of the database is in use, and whether anything needs attention.
86%
Share of reads answered from memory instead of disk.
Healthy databases sit above 99%. Below that, everyday queries wait on disk because the data no longer fits in memory.
25,038 GB
Data written to temporary files because a query outgrew its memory.
Sorting and grouping happen in memory until they no longer fit, then spill to disk and slow down sharply.
This snapshot covers capacity and overall health. Signing in with your @thebakersdozen.in account adds every table by name and size, the largest-tables breakdown, and the sortable detail view.
This page rebuilds itself. A scheduled job reads the database every six hours and redeploys, so nobody has to run anything by hand. Sizes come from the Postgres catalog and are true whenever you look; the health signals are counters that reset when the server restarts, which is why each one states the window it covers.