Schedule - PGConf.EU 2019
Caching long running queries
Date: 2019-10-16
Time: 14:40–15:30
Room: Foscolo
Level: Intermediate
There are various approaches to cache query results using various algorithms, such as generic LeastRecentlyUsed and TimeExpiry algorithms, generic low level caching with deltas (Nagata), or high level entity caching architecture using logical decoding (Atanasovski). Our time stamp deltas based approach sits between the low level algorithms and the high level architectural solution. It is developed to address a situation in our document storage system, where - the use of the queries is too irregular and often sporadic (ruling out LRU or TE) - no IDs are present - the system is not write intensive ( eg. 0-4 entries per second per relevance set) - the queries are complex - time stamp is a part of the query
Slides
The following slides have been made available for this session: