X-Received: by 10.224.127.70 with SMTP id f6mr5525020qas.5.1377029891362; Tue, 20 Aug 2013 13:18:11 -0700 (PDT) X-Received: by 10.50.138.138 with SMTP id qq10mr734818igb.3.1377029891056; Tue, 20 Aug 2013 13:18:11 -0700 (PDT) Path: news.cambrium.nl!textnews.cambrium.nl!feeder2.cambriumusenet.nl!feeder1.cambriumusenet.nl!feed.tweaknews.nl!209.85.216.87.MISMATCH!f7no3625548qan.0!news-out.google.com!he10ni2777qab.0!nntp.google.com!fx3no3787051qab.0!postnews.google.com!glegroupsg2000goo.googlegroups.com!not-for-mail Newsgroups: comp.databases.oracle.server Date: Tue, 20 Aug 2013 13:18:10 -0700 (PDT) In-Reply-To: <98794f2c-ace0-4567-bad3-ee1a4993a9d5@googlegroups.com> Complaints-To: groups-abuse@google.com Injection-Info: glegroupsg2000goo.googlegroups.com; posting-host=64.58.150.28; posting-account=tpQovAkAAABNoH5bwsZAiff2L0zxGwdv NNTP-Posting-Host: 64.58.150.28 References: <98794f2c-ace0-4567-bad3-ee1a4993a9d5@googlegroups.com> User-Agent: G2/1.0 MIME-Version: 1.0 Message-ID: Subject: Re: Thoughts on instance crash and RAC From: joel garry Injection-Date: Tue, 20 Aug 2013 20:18:11 +0000 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: quoted-printable Xref: news.cambrium.nl On Tuesday, August 20, 2013 4:48:13 AM UTC-7, vsevolod afanassiev wrote: > Some thoughts on instance crash and RAC: >=20 > - Main selling point of RAC is that it increases availability: if one nod= e crashes other nodes are still available. During presentations Oracle guys= would crash an instance by killing PMON and show that the database is stil= l accessible through other nodes. >=20 > - However in my experience this isn't very realistic scenario. Oracle ins= tances rarely experience such =93instant=94 crash =96 an instance was runni= ng without problems and then few seconds later it is completely gone. More = often an instance would struggle for a while =96 may be for 10 minutes, may= be for half an hour. Things like latch waits, library cache wait, etc. wou= ld appear, may be with ORA-00600 and 7445 errors, may be 4031. However cras= h is unlikely or it will take a while. Instance "death" became slow and pai= nful. >=20 You should probably google: you probably don't need rac jg --=20 @home.com is bogus. http://www.zdnet.com/oracle-survey-examines-evolving-cfo-role-cooperation-w= ith-cios-7000019640/