Re: Using RSYNC for replication?
От | Vivek Khera |
---|---|
Тема | Re: Using RSYNC for replication? |
Дата | |
Msg-id | x7znplf9vk.fsf@onceler.kciLink.com обсуждение исходный текст |
Ответ на | Re: Using RSYNC for replication? (Tom Lane <tgl@sss.pgh.pa.us>) |
Список | pgsql-general |
>>>>> "TL" == Tom Lane <tgl@sss.pgh.pa.us> writes: TL> A final note is that I doubt this would be very efficient: wouldn't TL> rsync have to ship entire table files (and entire WAL log files) for TL> even the most piddling change? No, rsync is smart that way. It checksums hunks of the files, and sends only the parts that change. I did a test and the first rsync took me 4 hours. The second one the next day took about 1.5 hours. I was moving the data to another box, and the time to dump/restore/analyze the tables was enormous. Using rsync made it faster. First I did the rsync live (4 hours), then shut down the source postmaster, did the rsync again (only 1 hour) and brought up the new server. Downtime was 1 hour rather than 8. However for backup purposes, it makes no sense. -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Vivek Khera, Ph.D. Khera Communications, Inc. Internet: khera@kciLink.com Rockville, MD +1-240-453-8497 AIM: vivekkhera Y!: vivek_khera http://www.khera.org/~vivek/
В списке pgsql-general по дате отправления: