So I tested this when I got in the shop this morning. The memory database is roughly 450:1 over an sqLite database on the LAN. That is incredibly significant, if the goal is to start with a certain dataset, then query it over and over.
Bob S > On Apr 21, 2022, at 19:52 , Bob Sneidar via use-livecode > <use-livecode@lists.runrev.com> wrote: > > Hi all. > > Just an exercise in semi-futility here. I was curious about the actual > difference in performance between querying a memory based sqLite database, as > opposed to querying a mySQL database over a slow (iPhone hotspot) connection. > I expected it to be significant, but HOW significant I had no idea. > > Querying for one record produced a performance ratio of 2600 to 1. Some will > say, "Of course! What did you expect?" That is just the point. Many people > assume repeated queries of an SQL database over a network to be tolerable, > even if the connection is remote. But what if your application needs to make > repeated queries? At some point, the overhead of repeated queries will exceed > that of querying for all or a subset of the data once, storing it in a memory > database, then performing your repeated queries on the memory database > instead. > > I will test this over a LAN connection when I get into work tomorrow, but I > expect it to be, if not AS significant, certainly more that one would expect. > > Just food for thought. > > Bob S > > > _______________________________________________ > use-livecode mailing list > use-livecode@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your subscription > preferences: > http://lists.runrev.com/mailman/listinfo/use-livecode _______________________________________________ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-livecode
