Hello, The system I am working on will be composed of 1 central server where the H2 database will run and 100 servers that will connect via TCP/IP and read data from the H2 database found on the central server. Note that the 100 local servers will just read data, they will not write to the database.
We want to read all "entities" X from the database. There is a table X in the database. Each x has many y parts, which are kept in table Y. Each y has some z, w, u, t, r, v parts which are kept in tables Z, W, U, T, R, V. Tables Z, W, U, T, R, V are the child tables and reference Y through a foreign key. Table Y is the child table of X and references X through a foreign key. I think you got a point on how the database hierarchy looks like. I started to make some experiments and these are my findings so far. When we have 1 local server that reads the "entities" X from the database, the average time to read one entity x is shorter than when we have 10 local servers reading the entity. It looks the more local servers we have, the slower it gets on average. So far I ran these tests only on my laptops and I could run them only with maximum 20 (simulated) local servers. I am wondering if I should expect any scalability issues when 100 local servers will be in use, all reading the same info in the same time from the same database? Looking at the way the reading time seems to increase on average it looks like it could be possible that reading will take a much longer time when 100 servers are reading the db compared to when only 1 is reading. Taking much longer to read from the db when we'll use 100 servers will have a negative impact of the overall application. This increase does not seem to be linear, so I don't feel confident to make any estimations on how it will behave. Do you have any suggestions/experience on running h2 like this? Thanks a lot! PS: This is the db url we are using: jdbc.url=jdbc:h2:tcp://192.168.200.49:77777/myH2DB;MVCC=TRUE;DB_CLOSE_ON_EXIT=FALSE; WRITE_DELAY=15000;CACHE_SIZE=65536;MAX_MEMORY_ROWS=5000 Would using the MULTI_THREADED option improve the reading times when multiple servers are reading from the database? -- You received this message because you are subscribed to the Google Groups "H2 Database" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/h2-database. For more options, visit https://groups.google.com/d/optout.
