Received: (qmail 7593 invoked from network); 10 Oct 2009 04:18:13 -0500 Received: from freelists-180.iquest.net (HELO turing.freelists.org) (206.53.239.180) by static-ip-85-25-126-90.inaddr.intergenia.de with SMTP; 10 Oct 2009 04:18:11 -0500 Received: from localhost (localhost [127.0.0.1]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id 2C88ECB2CE6; Sat, 10 Oct 2009 05:17:26 -0400 (EDT) X-Virus-Scanned: Debian amavisd-new at localhost.localdomain Received: from turing.freelists.org ([127.0.0.1]) by localhost (turing.freelists.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id SBjYeL0LYaH3; Sat, 10 Oct 2009 05:17:25 -0400 (EDT) Received: from turing.freelists.org (localhost [127.0.0.1]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id 48F67CB2C86; Sat, 10 Oct 2009 05:16:48 -0400 (EDT) Received: with ECARTIS (v1.0.0; list oracle-l); Sat, 10 Oct 2009 05:16:06 -0400 (EDT) Received: from localhost (localhost [127.0.0.1]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id 7A7CDCB29A0 for ; Sat, 10 Oct 2009 05:16:06 -0400 (EDT) Received: from turing.freelists.org ([127.0.0.1]) by localhost (turing.freelists.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id lJL5Uuk7aCGi for ; Sat, 10 Oct 2009 05:16:06 -0400 (EDT) Received: from mail-fx0-f221.google.com (mail-fx0-f221.google.com [209.85.220.221]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id C3AEFCB1815 for ; Sat, 10 Oct 2009 05:16:05 -0400 (EDT) Received: by fxm21 with SMTP id 21so6310436fxm.33 for ; Sat, 10 Oct 2009 02:16:04 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:content-type; bh=mw9B1e6bfPI3361GKEelCzQOeoRxmAvfColXih5QUOo=; b=ASdcJmjsx5y7vGQNPejELBd7+W094KfWEt2tXYFVxK3NXnGfI/5FfO2GdN6rEwddAM 01UEATdEiX6gL0jLawXA9qT14+L44m/cluIERRnZw79GUudm+ceJb0fDnV5s79rpzgH1 pUH0pMMjap/BwG8sHiBDnb6f4d9I7/GPTQrAc= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=Psplsqhn9SqOOsEKytv6jCq8CIEQDrBOArKGz+tDEmlAF8skR39TUREPgCM4BXDRy9 AdOHPHyLsfNmmz7zZQzFzswF7YcEHMa6/3sNPgb0aB6ZqlR48B+zXJPEwHRacftsVX5V 1rb88jDvRBI+rQ6xoIIEFDns7IhB2CkUQWcyc= MIME-Version: 1.0 Received: by 10.223.58.139 with SMTP id g11mr1045842fah.43.1255166164807; Sat, 10 Oct 2009 02:16:04 -0700 (PDT) In-Reply-To: <713d96d10910100151v2047e692s9d7b2d3960945093@mail.gmail.com> References: <4acf6ff5.0ec5100a.5170.4bd2@mx.google.com> <34f494ce0910091057h52d79732j324d9faf6da5c774@mail.gmail.com> <5e3048620910092058p5788cd6aq3b69acf745799a52@mail.gmail.com> <713d96d10910100151v2047e692s9d7b2d3960945093@mail.gmail.com> Date: Sat, 10 Oct 2009 11:16:04 +0200 Message-ID: Subject: Re: major blunders From: Frits Hoogland To: oracle-l@freelists.org Content-Type: multipart/alternative; boundary=000e0ce0eea86331ae04759125a4 X-archive-position: 21519 X-ecartis-version: Ecartis v1.0.0 Sender: oracle-l-bounce@freelists.org Errors-to: oracle-l-bounce@freelists.org X-original-sender: frits.hoogland@gmail.com Precedence: normal Reply-to: frits.hoogland@gmail.com List-help: List-unsubscribe: List-software: Ecartis version 1.0.0 List-Id: oracle-l X-List-ID: oracle-l List-subscribe: List-owner: List-post: List-archive: X-list: oracle-l --000e0ce0eea86331ae04759125a4 Content-Type: text/plain; charset=ISO-8859-1 not really major, still happening: the unix sysadmin who decides to clean up those pesky large logfiles on which process still have a file handle open. files gone (as seen from the sysadmin's session) but space not reclaimed... frits On Sat, Oct 10, 2009 at 10:51 AM, Howard Latham wrote: > Rman. > > If you use rman have 2 batch jobs for each backup. > One that does crosschecks deletes etc > then a Second one that actually does the backup. > that way you can search the second log for RMAN- errors and detect when a > backup has really failed not when one > backup piece from months ago is missing or some other - while significant > not a major priority to fix. > > > > > 2009/10/10 Connor McDonald > > >> >> On Sat, Oct 10, 2009 at 2:19 AM, Jared Still wrote: >> >>> Every little bit helps. >>> Using rm * has already been dismissed as a bad idea, >>> so it would seem that rm somedir/* would also be a questionable >>> practice. >>> >>> If there's a way to screw up, all of will find it at some point. >>> >>> The idea is to mitigate the risk as much as possible. >>> >>> >>> Jared Still >>> Certifiable Oracle DBA and Part Time Perl Evangelist >>> Oracle Blog: http://jkstill.blogspot.com >>> Home Page: http://jaredstill.com >>> >>> >>> >>> On Fri, Oct 9, 2009 at 10:57 AM, Aaron Leonard wrote: >>> >>>> I'd be very wary of relying on that. It doesn't account for when you >>>> are not in the directory which contain the files your are deleting. >>>> >>>> oracle:~/adl/tmp> for i in 1 2 3 4 5 6 7; do echo 1 > file$i; done >>>> oracle:~/adl/tmp> touch -- -i >>>> oracle:~/adl/tmp> rm -f * >>>> rm: remove regular file `file1'? >>>> oracle:~/adl/tmp> rm -f * >>>> rm: remove regular file `file1'? n >>>> rm: remove regular file `file2'? n >>>> rm: remove regular file `file3'? n >>>> rm: remove regular file `file4'? n >>>> rm: remove regular file `file5'? n >>>> rm: remove regular file `file6'? n >>>> rm: remove regular file `file7'? n >>>> oracle:~/adl/tmp> n >>>> bash: n: command not found >>>> oracle:~/adl/tmp> cd .. >>>> oracle:~/adl> rm -f tmp/* >>>> oracle:~/adl> ls -l tmp >>>> total 0 >>>> oracle:~/adl> >>>> >>>> >>>> >>>> On Fri, Oct 9, 2009 at 12:44 PM, Jared Still wrote: >>>> >>>>> On Fri, Oct 9, 2009 at 10:16 AM, Fmhabash wrote: >>>>> >>>>>> If I may add ... >>>>>> - I altered behavior of 'rm' cmd forcing it to be interactive needing >>>>>> user confirmation. Scripts coded to use non-interactive version. >>>>>> >>>>> >>>>> A trick to avoid accidentally deleting files you would rather keep. >>>>> >>>>> Create a file called '-i' in key directories. This will force rm -f to >>>>> go into interactive >>>>> mode, as the '-i' filename is interpreted as an argument. >>>>> >>>>> Create '-i' file: touch -- -i >>>>> >>>>> The double dash is used to tell rm (or any other *nix command) that >>>>> the >>>>> following stuff on the command line is not a command line option. >>>>> >>>>> Remove the file: rm -- -i >>>>> >>>>> Create a tmp directory, copy a few files into it, and try it. >>>>> >>>>> It has saved me at least once. >>>>> >>>>> >>>>> Jared Still >>>>> Certifiable Oracle DBA and Part Time Perl Evangelist >>>>> Oracle Blog: http://jkstill.blogspot.com >>>>> Home Page: http://jaredstill.com >>>>> >>>>> >>>> >>> >> >> I've already posted this in the past, but I love the approach taken at a >> client I once worked at. All production servers had a naming scheme of: >> bfc"digits" >> When I asked what "bfc" stood for... it was "be f...ing careful". They >> had adopted it in all seriousness because it >> >> a) made you pause for a chuckle every time you saw it >> b) that pause impressed the seriousness of the situation to you >> >> >> >> -- >> Connor McDonald >> =========================== >> email: connor_mcdonald@yahoo.com >> web: http://www.oracledba.co.uk >> >> "Semper in excremento, sole profundum qui variat" >> > > > > -- > Howard A. Latham > > > --000e0ce0eea86331ae04759125a4 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable not really major, still happening:

the unix sysadmin who= decides to clean up those pesky large logfiles on which process still have= a file handle open. files gone (as seen from the sysadmin's session) b= ut space not reclaimed...

frits

On Sat, Oct 10,= 2009 at 10:51 AM, Howard Latham <howard.latham@gmail.com> wrote:
Rman.

If you use rman have 2 batch jobs for each backup.
One that= does crosschecks deletes etc
then a Second one that actually does the b= ackup.
that way you can search the second log for RMAN- errors and detec= t when a backup has really failed not when one
backup piece from months ago is missing or some other - while significant n= ot a major priority to fix.




2= 009/10/10 Connor McDonald <mcdonald.connor@gmail.com>


On Sat, Oct 10, 2009 at 2:19 AM, Jared Sti= ll <jkstill@gmail.com> wrote:
Every little bit helps.

Using rm * has already been dism= issed as a bad idea,
so it would seem that rm somedir/* would als= o be a questionable
practice.

If there&#= 39;s a way to screw up, all of will find it at some point.

The idea is to mitigate the risk as much as possible.


Jared Still
Certifiable O= racle DBA and Part Time Perl Evangelist
Oracle Blog: http://jkstill.blogspot.com
Home Page: http://jared= still.com



On Fri, Oct 9= , 2009 at 10:57 AM, Aaron Leonard <aachleon@gmail.com> wrot= e:
I'd be very wary of rely= ing on that.=A0 It doesn't account for when you are not in the director= y which contain the files your are deleting.

oracle:~/adl/tmp> for i in 1 2 3 4 5 6 7; do echo 1 > file$i;= done
oracle:~/adl/tmp> touc= h -- -i
oracle:~/adl/tmp> rm -f *<= br style=3D"font-family:courier new,monospace"> rm: remove regular file `= file1'?
oracle:~/adl/tmp> rm -f *
rm: remove regular file `= file1'? n
rm: remove regular file `file2&#= 39;? n
rm: remove regular file `= file3'? n
rm: remove regular file `file4&#= 39;? n
rm: remove regular file `= file5'? n
rm: remove regular file `file6&#= 39;? n
rm: remove regular file `= file7'? n
oracle:~/adl/tmp> n bash: n: command not foun= d
oracle:~/adl/tmp> cd ..
oracle:~/adl> rm -f tm= p/*
oracle:~/adl> ls -l tmp
total 0
oracle:~/adl>



On Fri, Oct 9, 2009 at 12:44 PM, Jared S= till <jkstill@gmail.com> wrote:
On Fri, Oct 9, 2009 at 10:16 AM, Fmhabash <fmh= abash@gmail.com> wrote:
If I may add ...
=A0- I altered behavior of 'rm' cmd forcing it = to be interactive needing user confirmation. Scripts coded to use non-inter= active version.
=A0
A trick to avoid acciden= tally deleting files you would rather keep.

Create a file called '-i' in key directories. = =A0This will force rm -f to go into interactive
mode, as the '= ;-i' filename is interpreted as an argument.

Create '-i' file: =A0touch -- -i

The doubl= e dash is used to tell rm (or any other *nix command) that the=A0
following stuff on the command line is not a command line option.

Remove the file: =A0rm -- -i

Create= a tmp directory, copy a few files into it, and try it.

It has saved me at least once.


Jared Still
Certifiable Oracle DBA and Part Time Perl Evangelist
Orac= le Blog: http://j= kstill.blogspot.com
Home Page: http://jaredstill.com





I've already poste= d this in the past, but I love the approach taken at a client I once worked= at. =A0All production servers had a naming scheme of: bfc"digits"= ;

When I asked what "bfc" stood for... it was "be f...ing= careful". =A0They had adopted it in all seriousness because it
<= div>
a) made you pause for a chuckle every time you saw it
b) that pause impressed the seriousness of the situation to you
<= div>


-- =
Connor McDonald
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
email: connor_mcdonald@yahoo.com
web: =A0 http://ww= w.oracledba.co.uk

"Semper in excremento, sole profundum qui= variat"



--
Howard A. Latham



--000e0ce0eea86331ae04759125a4-- -- http://www.freelists.org/webpage/oracle-l