[h2] JDBC getIndexInfo for large table is very slow

2022-09-30 Thread Silvio
I have a largisch database (~6G) which I use in H2 2.1.214 embedded mode on an NVME SSD. I use the JDBC call rsIdx = con.getMetaData.getIndexInfo(catalog,schema,table,false,false) to get the indexes for a single table (which is by far the largest in the whole database) that contains about 3.6 m

Re: [h2] JDBC getIndexInfo for large table is very slow

2022-09-30 Thread Noel Grandin
How many tables do you have, that should return in milliseconds. You could try rinning the built in profiler (see the docs) and see what it is doing. You could also directly query the metadata table INFORMATION_SCHEMA.INDEXES -- You received this message because you are subscribed to the Google

Re: [h2] JDBC getIndexInfo for large table is very slow

2022-09-30 Thread Silvio
Thanks Noel, Also see my other remark. I will look into that table. Actually, we did some stuff directly on the INFORMATION_SCHEMA tables before but the H2 versions above 200 had some modified layout/content in there so we switched to the JDBC call for this one. On Friday, 30 September 2022 at

Re: [h2] JDBC getIndexInfo for large table is very slow

2022-09-30 Thread Silvio
BTW: there are usually about 30 tables. Some approaching 100K rows, the big one being the only that goes into the millions. On Friday, 30 September 2022 at 12:52:49 UTC+2 Silvio wrote: > Thanks Noel, > > Also see my other remark. I will look into that table. Actually, we did > some stuff direct