[h2] insertion after drop is much slower than insertion in a new Table

Hello, I am inserting (with batch insert - bulk=1) around 350,000 rows in a table, when the table (and File DB) does not exist the load is very fast, then I rerun the test which drop the table before (if exists) then commit then recreate the table, the insertion then is very slow and there

Re: [h2] insertion after drop is much slower than insertion in a new Table

I rerun the test with trace_level=2, with level=3 the file was extremely big but if needed, I can try to do that also with another set of data, but my results with trace_level 2 is: First Run, the database does not yet exist: - -- SQL Statement Statistics

Re: [h2] insertion after drop is much slower than insertion in a new Table

I did this, I replaced application code with statement.executeBatch() // which should insert 10,000 inserts at once, the result is this: H2: Profiler: top 3 stack trace(s) of of 254 ms of 127 thread dumps: 120/126 (95%): at org.h2.store.fs.FileDisk.force(FilePathDisk.java:410) at org.h2.store.Fil

Re: [h2] insertion after drop is much slower than insertion in a new Table

; > On 2014-04-07 15:35, Kais Haddadin wrote: > >> >> H2: Profiler: top 3 stack trace(s) of of 254 ms of 127 thread dumps: >> 120/126 (95%): >> at org.h2.store.fs.FileDisk.force(FilePathDisk.java:410) >> at org.h2.store.FileStore.sync(FileStore.java:419) >>

Re: [h2] insertion after drop is much slower than insertion in a new Table

thinking about this again I don't understand why this should make a difference if the table is already there or not, if 32 is a small value then it should be slow when I insert the 366328 in the first run as well, right? Cheers, Kais On Mon, Apr 7, 2014 at 4:29 PM, Kais Haddadin wrote: &

Re: [h2] Re: insertion after drop is much slower than insertion in a new Table

It will take some time to do that I will try to do that as soon as possible and post it here. Cheers, Kais -- 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 h2-da

Re: [h2] Re: insertion after drop is much slower than insertion in a new Table

with a separate example (attached) I can not reproduce the problem. From what I saw from the trace file, the attached test should be very similar. I will further investigate from my application. Regards, Kais On Tue, Apr 8, 2014 at 8:58 AM, Kais Haddadin wrote: > It will take some time to

Re: [h2] Re: insertion after drop is much slower than insertion in a new Table

Hi Thomas, I will try out the commit every n rows, I think it will solve my problem too. But about the parameters mentioned in the link above: LOG=0;CACHE_SIZE=65536;LOCK_MODE=0;UNDO_LOG=0 I used to use them before, but I got the problem that the database gets corrupted in some unknown cases. As I

Re: [h2] Re: insertion after drop is much slower than insertion in a new Table

-09 09:18, Kais Haddadin wrote: > >> I will try out the commit every n rows, I think it will solve my problem >> too. But about the parameters mentioned in the >> link above: >> LOG=0;CACHE_SIZE=65536;LOCK_MODE=0;UNDO_LOG=0 >> I used to use them before, but I got the p

Re: database corruption issue

sorry I posted this twice, this one can be closed On Wednesday, October 17, 2012 10:55:25 AM UTC+2, kais haddadin wrote: > > Hello! > > when I am trying to delete a table, I got this error: > > Allgemeiner Fehler: "java.lang.ArrayIndexOutOfBoundsExc

Re: Database corruption problem

13' //this gives an error On Wednesday, October 17, 2012 10:46:29 AM UTC+2, kais haddadin wrote: > > Hello! > > I am trying to run a simple delete on a table. This error comes up: > > Allgemeiner Fehler: "java.lang.ArrayIndexOutOfBoundsException" > General er

is there a limit of how many records in one table?

Hello! I am trying to load big table: 250 columns and 6 million lines in H2. It loads (or I think at least it is because the db is now 11.5GB), but It gives me always out of memory as soon as I try to make a query on it. I stopped my program andI tried to access the H2 database outside my own

Re: is there a limit of how many records in one table?

hread > dumps? > > How exactly did you load the database? You wrote "I am trying to load big > table" but when I read your mail I'm not sure: did you finish loading the > database, and now try to access it (read from it)? > > By the way, according to the error code

Re: is there a limit of how many records in one table?

On Mon, Nov 5, 2012 at 11:02 AM, Kais Haddadin wrote: > > Hi, > > sorry about this, but I think there was some confusions about the versions > of the H2 used in my test. It seems that there was older versions involved. > I redid the test completely with version 169 and I co

Re: is there a limit of how many records in one table?

, 2012 at 12:58 PM, Noel Grandin wrote: > Nice work. > I think you have something wrong with your query, because it's trying to > return a seriously large number of records. > > > On 2012-11-05 13:54, kais haddadin wrote: > > Ok I did what you recommended. The dum

Problem opening H2

Hello, I try to open the H2 database version (1.3.168) and am getting this error: "Sequenz "SYSTEM_SEQUENCE_A5042270_9313_42F1_B53B_FC2AD0D3510F" nicht gefunden Sequence "SYSTEM_SEQUENCE_A5042270_9313_42F1_B53B_FC2AD0D3510F" not found; SQL statement: CREATE CACHED TABLE FCM.LAGERBESTÄNDE( "

Re: Problem opening H2

the stacktrace is attached On Wed, Jan 9, 2013 at 2:13 PM, Noel Grandin wrote: > Can you please post the entire stacktrace? > > > On 2013-01-09 15:03, kais haddadin wrote: > > Hello, > > I try to open the H2 database version (1.3.168) and am getting thi

Re: Problem opening H2

Ok, I get it, then LOG=0 is not recommended for productive system. Sorry about this, I already had problems because I used LOCK_MODE=0. But it was not clear for me that these parameters also destabilize the database. But what about UNDO_LOG, if I do not have roll_back operation in the code, is it s