Обсуждение: Re: [9.0] hot standby plus streaming replication
Ciao Michele, > both server (master and standby) need a common directory where read and write the wal files? Not necessarily. You can use for instance scp to ship the WAL file from the master to the standby using the network. > Just another question about replication: there is the possibility to build a sync between a 32 and a 64 bit (on linux)? As stated in the documentation (http://www.postgresql.org/docs/9.0/static/warm-standby.html), the hardware architecture must be the same. Hope this helps. Ciao, Gabriele -- Gabriele Bartolini - 2ndQuadrant Italia PostgreSQL Training, Services and Support gabriele.bartolini@2ndQuadrant.it | www.2ndQuadrant.it
Gabriele Bartolini ha scritto: > Ciao Michele, Ciao ;) > >> both server (master and standby) need a common directory where >> read >> > and write the wal files? > > Not necessarily. You can use for instance scp to ship the WAL file > from the master to the standby using the network. > Thanks for the explain, but... - why in my tests, _whitout_ common direcotory, master and slave keep in sync also if I shutdown slave, add (in my last tests) something about 100k record (although little ones) on the master and then after woke up the slave in about 2/3 seconds I have all the dbs in sync? It's just luck or there are something that I don't understand? - If I have to copy (scp / rsync / etc...) the files, which interval I have to follow? And also, when the slave will read these files? There are something like inotify or I have to say to slave "my friend, there are updates from master. Keep in sync!" >> Just another question about replication: there is the possibility >> to > build a sync between a 32 and a 64 bit (on linux)? > > As stated in the documentation > (http://www.postgresql.org/docs/9.0/static/warm-standby.html), the > hardware architecture must be the same. > If I read before this page, I spared a lot of tries! > Hope this helps. > > Ciao, Gabriele > Thanks a lot, Michele P.s. Glad to see that also in Italy there are PostgreSQL guru ;)
> > P.s. Glad to see that also in Italy there are PostgreSQL guru ;) > Glad to see that more people are using Indian words (Guru) :) http://en.wikipedia.org/wiki/Guru Regards, Jayadevan DISCLAIMER: "The information in this e-mail and any attachment is intended only for the person to whom it is addressed and may contain confidential and/or privileged material. If you have received this e-mail in error, kindly contact the sender and destroy all copies of the original communication. IBS makes no warranty, express or implied, nor guarantees the accuracy, adequacy or completeness of the information contained in this email or any attachment and is not liable for any errors, defects, omissions, viruses or for resultant loss or damage, if any, direct or indirect."
On Fri, Oct 1, 2010 at 4:02 PM, Michele Petrazzo - Unipex <michele.petrazzo@unipex.it> wrote: > - why in my tests, _whitout_ common direcotory, master and slave keep in > sync also if I shutdown slave, add (in my last tests) something about > 100k record (although little ones) on the master and then after woke up > the slave in about 2/3 seconds I have all the dbs in sync? Because the master had the WAL files containing that 100k record in its pg_xlog directory. If those WAL files were unfortunately removed from the master before you started the standby, the standby would not have been in sync with the master. You can specify how many WAL files you'd keep in the master by using wal_keep_segments parameter. http://developer.postgresql.org/pgdocs/postgres/runtime-config-wal.html#RUNTIME-CONFIG-REPLICATION Regards, -- Fujii Masao NIPPON TELEGRAPH AND TELEPHONE CORPORATION NTT Open Source Software Center
Fujii Masao ha scritto: > On Fri, Oct 1, 2010 at 4:02 PM, Michele Petrazzo - Unipex > <michele.petrazzo@unipex.it> wrote: >> - why in my tests, _whitout_ common direcotory, master and slave >> keep in sync also if I shutdown slave, add (in my last tests) >> something about 100k record (although little ones) on the master >> and then after woke up the slave in about 2/3 seconds I have all >> the dbs in sync? > > Because the master had the WAL files containing that 100k record in > its pg_xlog directory. If those WAL files were unfortunately removed > from the master before you started the standby, the standby would > not have been in sync with the master. > This was the explain that I was looking for! Do you know if there is a talk or a "best practice" about keep in sync a master/slave couple without a shared directory? Something that talk about the number of a typical data that can be keep in sync. Can be very useful and more simple to maintain an installation without a shared than a one with it and I think that more users will be happy with! > You can specify how many WAL files you'd keep in the master by using > wal_keep_segments parameter. > http://developer.postgresql.org/pgdocs/postgres/runtime-config-wal.html#RUNTIME-CONFIG-REPLICATION > Seen! Thanks a lot, Michele