Hi list,
I have executed a lot of "LOAD GRAPH" and i have a problem with the biggest.
i loaded 5 graph.
* 200.000
* 1.900.000
* 13.000.000
* 38.000.000
* +2.000.000.000
For the 4 littlest, i have no problem.
For the last (uniprot), the LOAD stopped without any message.
The procedure to LOAD a GRAPH:
isql-vt ....
set autocommit on;
DB.DBA.ld_dir ('/data/import/uniprot','uniprot.rdf.gz','http://uniprot');
DB.DBA.rdf_loader_run ();
And at the moment, it work.
No problem of RAM or SWAP (thansk for the autocommit), no problem of space
left in the hard drive, no problem of connection.
- The DB file size grows up to 40GB.
- The DB last modified is 6h36am the next day.
- The log message are empty or almost.
09:56:55 INFO: HTTP/WebDAV server online at 8890
09:56:55 INFO: Server online at 1111 (pid 5673)
09:58:43 INFO: PL LOG: Loader started
21:57:06 INFO: Checkpoint started
21:57:53 INFO: Checkpoint finished, new log is
/data/virtuoso/xlog/virtuoso20110809215656.trx
- The server is UP but it's impossible to connect it (isql-vt)
tcp 0 0 *:1111 *:*
LISTEN
5673 ? Ssl 1039:17 /usr/bin/virtuoso-t +wait +configfile
/conf/virtuoso/virtuoso.ini +debug
After a restart of the processus, i execute a "COUNT" on the graph and there
are 412.540.301 triplets loaded. Far from 2 billions.
I doesn't lost the terminal, all indicate the "LOAD" continu except the
CPU, the DB file size, the DB file time, and the iowait.
In another "LOAD", i have a graph up to 1.700.000.000 triplets and a DB file
size of 98GB. But for an inexplicable reason, the "LOAD" was interrupt.
Thanks.
Maxime T.
