Thanks, but I need to do this programmatically, not via the shell. The tables aren't huge - they only contain test data that needs to be cleaned up after the test cycle completes.
Ishaaq stack-3 wrote: > > You could do that but would probably take some time. Or you could > truncate > the table via the shell. This does disable, drop, recreate in 0.19.x. In > TRUNK, it does disable, drop, major compaction of .META. table, and then > recreate (there is a known issue around dropping a big table. A > workaround > is to run major compaction on the .META. table. In 0.19.x, you can do > this > from the shell by doing 'major_compact ".META."'). Be warned that if your > table is big, users report our enable/disable is a flakey. You might have > to do it a few times to ensure all regions disabled before delete will > succeed (fix coming...). > > St.Ack > > > On Tue, Jul 7, 2009 at 10:34 PM, Ishaaq Chandy <[email protected]> wrote: > >> >> Hi guys, >> What is the quickest way to delete all the rows in a table without >> deleting >> the table itself? The only way I can see would be to use a scanner. Is >> this >> right? I am using 0.19.3. >> >> Thanks, >> Ishaaq >> -- >> View this message in context: >> http://www.nabble.com/delete-all-rows-in-table-tp24385603p24385603.html >> Sent from the HBase User mailing list archive at Nabble.com. >> >> > > -- View this message in context: http://www.nabble.com/delete-all-rows-in-table-tp24385603p24385875.html Sent from the HBase User mailing list archive at Nabble.com.
