Postgresql: how an index is physically created on disk
I am interested in where the files (holding the index) are being written, instead of the various indexing algorithms involved in building an index.
Read more…
I am interested in where the files (holding the index) are being written, instead of the various indexing algorithms involved in building an index.
Read more…
The contained database is named “test”. It contains only one table, whose schema is
create table tb(f1 int, f2 int);
Suppose your postgresql server runs on machine h1, and you want to connect to it from machine h2. So, within a console of h2, you run the following command
psql -h h1 mydb
When testing with DBT2, the driver needs to open 2 files for every process, especially when you launch client and server on the same machine.
Read more…
It’s simple once you know where to look:
Read more…
Suppose you have multiple lines of numbers in a file called “numbers”.
Read more…
sudo blockdev –getsize64 /dev/sd? | awk ‘{print($1/1024/1024/1024);}’
Of course, you will need to do “ls /dev/sd?” to have an idea of which number corresponds to which device.
I met this error message when running DBT2 script “./scripts/run_workload.sh” against Postgresql.
Read more…
Like the previous two posts, this is again a problem I met when generating data into Postgresql with DBT2. The full error message is:
Read more…
Recent Comments