Received: from localhost (unknown [200.46.204.184]) by postgresql.org (Postfix) with ESMTP id 137B32E011F for ; Tue, 4 Dec 2007 07:19:21 -0400 (AST) Received: from postgresql.org ([200.46.204.71]) by localhost (mx1.hub.org [200.46.204.184]) (amavisd-maia, port 10024) with ESMTP id 41761-06 for ; Tue, 4 Dec 2007 07:19:14 -0400 (AST) Received: from news.hub.org (news.hub.org [200.46.204.72]) by postgresql.org (Postfix) with ESMTP id 77EC02E0049 for ; Tue, 4 Dec 2007 07:19:20 -0400 (AST) Received: from news.hub.org (news.hub.org [200.46.204.72]) by news.hub.org (8.14.1/8.14.1) with ESMTP id lB4BJ9Ds092940 for ; Tue, 4 Dec 2007 07:19:10 -0400 (AST) (envelope-from news@news.hub.org) Received: (from news@localhost) by news.hub.org (8.14.1/8.14.1/Submit) id lB4B93W5089570 for pgsql-admin@postgresql.org; Tue, 4 Dec 2007 07:09:03 -0400 (AST) (envelope-from news) From: Aldor X-Newsgroups: pgsql.admin Subject: Re: Replication advice Many-to-one Slony Date: Tue, 04 Dec 2007 12:08:45 +0100 Organization: Hub.Org Networking Services Lines: 59 Message-ID: References: <20071018114038.M97749@biomag.usp.br> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit X-Complaints-To: usenet@news.hub.org User-Agent: Thunderbird 2.0.0.9 (Windows/20071031) In-Reply-To: <20071018114038.M97749@biomag.usp.br> To: pgsql-admin@postgresql.org X-Virus-Scanned: Maia Mailguard 1.0.1 X-Archive-Number: 200712/14 X-Sequence-Number: 27531 Hello Walfred, yes you can use slony, we use slony for huge data replication. One solution of them is the backup of many masters to one slave. It is in real time - and we're quite happy with the newest version of slony. Don't forget to not only add tables to your replication but to also add sequences additionally, because you may backup the tables and content of them but when trying to use your backup you could have a big nightmare when you cannot find the state of your sequences - so your data is not lost but not instantly usable. When you think of using a hot-copy system so that you just can switch from the master when it brakes to your slave system I would suggest you to think about the load that could go to the backup-slave which should now be used as master, in this case we calculate 1/3 of master-load to one slave, but that depends on the application and the load of your databases. Switching to Slave and back to master when it's again ready (also with automatic replication back to the master) works great in slony. If you have any further questions or need any help on setting up the slony scripts just contact me directly. Regards, Aldor Walfred Tedeschi wrote: > Hi all, > > We intend to establish a central back-up for our system. It can be though as > many machines (masters) linked to one Slave. The communication may be one > directional (master -> slave) and would appreciate to have it almost in realtime. > The data flux on the system is not very high in terms of transactions and > volume of data. And the primary keys are planned in a way that there is no > overlap over the masters (every master has its own range of PK) without > overlapping. We were thinking about slony1. > > Since we are not sure about the decision, we would appreciate very much if you > could advise us what to use and where to start. > > Thank you very much, > > Fred > > > -- > Open WebMail Project (http://openwebmail.org) > > > ---------------------------(end of broadcast)--------------------------- > TIP 1: if posting/reading through Usenet, please send an appropriate > subscribe-nomail command to majordomo@postgresql.org so that your > message can get through to the mailing list cleanly >