Received: from localhost (maia-2.hub.org [200.46.204.187]) by postgresql.org (Postfix) with ESMTP id 7A85A9FBC4B for ; Wed, 14 Feb 2007 03:20:19 -0400 (AST) Received: from postgresql.org ([200.46.204.71]) by localhost (mx1.hub.org [200.46.204.187]) (amavisd-new, port 10024) with ESMTP id 13678-01-6 for ; Wed, 14 Feb 2007 03:20:12 -0400 (AST) X-Greylist: delayed 00:16:26.932078 by SQLgrey-1.7.4 Received: from wp001.webpack.hosteurope.de (wp001.webpack.hosteurope.de [80.237.132.8]) by postgresql.org (Postfix) with ESMTP id 2ACE89FBC85 for ; Wed, 14 Feb 2007 03:11:48 -0400 (AST) Received: from p54ac3cfe.dip0.t-ipconnect.de ([84.172.60.254] helo=[192.168.1.113]); authenticated by wp001.webpack.hosteurope.de running ExIM using esmtpa id 1HHE2x-00038w-JU; Wed, 14 Feb 2007 07:55:19 +0100 Message-ID: <45D2B252.4010109@thopap.de> Date: Wed, 14 Feb 2007 07:55:14 +0100 From: Thomas Papke User-Agent: Mozilla Thunderbird 1.0.7 (Windows/20050923) X-Accept-Language: de-DE, de, en-us, en MIME-Version: 1.0 To: Peter Eisentraut CC: pgsql-admin@postgresql.org Subject: Re: Question to safe way for minor update References: <1001305.61140.62.214.193.211.1171388243.squirrel@webmailer.hosteurope.de> <200702132111.23707.peter_e@gmx.net> In-Reply-To: <200702132111.23707.peter_e@gmx.net> Content-Type: multipart/alternative; boundary="------------050400010008030804040109" X-bounce-key: webpack.hosteurope.de;tp@thopap.de;1171437108;6e559205; X-Virus-Scanned: Maia Mailguard 1.0.1 X-Spam-Status: No, hits=0.403 tagged_above=0 required=5 tests=BAYES_40, HTML_30_40, HTML_MESSAGE, HTML_TITLE_EMPTY, SPF_PASS X-Spam-Level: X-Archive-Number: 200702/190 X-Sequence-Number: 24655 This is a multi-part message in MIME format. --------------050400010008030804040109 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Right now our 8.1.4 is running without any compiler options. Any possible problem if i will compile the 8.1.8 with CLAGS -O2 -march=Opteron --> will the datafiles of the 8.1.4 non optimized postgres make any problems? Thanks a lot, Thomas Peter Eisentraut schrieb: >Thomas Papke wrote: > > >>1. i am running a large postgres 8.1.4 database in production. What >>is a good and safe way for an minor update to 8.1.8 (less downdown >>and possible jump back option)? >> >> > >Using binary packages, install the new one over the old one, and if you >don't like it, install the old one over the new one. > > > >>How riskful is an minor update for the data files? >> >> > >Not riskful. > > > >>2. i am using two dualcore opterons with 8GB memory - if i am using >>the CFLAGS="-O2" option for compile optimization --> is there any >>real performance boost? >> >> > >I don't think anyone has ever seriously considered building with >anything less than -O2 for anything but debugging purposes or coping >with buggy compilers on rare architectures. In other words, there >isn't any good reason not to do it. > > > --------------050400010008030804040109 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Right now our 8.1.4 is running without any compiler options. Any possible problem if i will compile the 8.1.8 with CLAGS -O2 -march=Opteron --> will the datafiles of the 8.1.4 non optimized postgres make any problems?

Thanks a lot,
Thomas

Peter Eisentraut schrieb:
Thomas Papke wrote:
  
1. i am running a large postgres 8.1.4 database in production. What
is a good and safe way for an minor update to 8.1.8 (less downdown
and possible jump back option)?
    

Using binary packages, install the new one over the old one, and if you 
don't like it, install the old one over the new one.

  
How riskful is an minor update for the data files?
    

Not riskful.

  
2. i am using two dualcore opterons with 8GB memory - if i am using
the CFLAGS="-O2" option for compile optimization --> is there any
real performance boost?
    

I don't think anyone has ever seriously considered building with 
anything less than -O2 for anything but debugging purposes or coping 
with buggy compilers on rare architectures.  In other words, there 
isn't any good reason not to do it.

  

--------------050400010008030804040109--