Networker

Re: [Networker] networker/oracle bug

2002-09-10 18:08:46
Subject: Re: [Networker] networker/oracle bug
From: Duncan Greenwood <duncan.greenwood AT BTINTERNET DOT COM>
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Date: Tue, 10 Sep 2002 22:58:36 +0100
According to the docs for the stat system call, a write should
update both mtime and ctime. I can't see how Oracle could
avoid this this happening, even intentionally.

st_atime
Time when file data was last accessed. Changed by the following
functions: creat(), mknod(), pipe(), utime(2), and read(2).

st_mtime
Time when data was last modified. Changed by the following
functions: creat(), mknod(), pipe(), utime(), and write(2).

st_ctime
Time when file status was last changed. Changed by the following
functions: chmod(), chown(), creat(), link(2), mknod(), pipe(), unlink(2),
utime(), and write().

There are some options to the NFS mount command that can
affect how these timestamps are handled. What is the full mount
command you are using, or the relevant line from /etc/vfstab
for the filesystem in question ?

D
#

----- Original Message -----
From: "Peter D. Gray" <pdg AT UOW.EDU DOT AU>
To: <NETWORKER AT LISTMAIL.TEMPLE DOT EDU>
Sent: Tuesday, September 10, 2002 7:58 AM
Subject: [Networker] networker/oracle bug


> I have a problem and I am not sure if
> it is networker or oracle or both.
>
> I am using snapshots to back up oracle databases
> on a procom filer.
>
> The problem is that the actual database files are not being
> picked up in incremental backups. The problem appears to
> be that oracle updates the atime and mtime timestamps
> when it writes to the database, but does not update the
> ctime.
>
> According to the man pages, both the ctime and mtime are
> updated when a write occurs to a file, so it is likely
> that oracle is playing dirty pool by not updating the
> ctime.
>
> However, it seems to me that networker is also
> playing dirty by not taking account of the mtime
> timestamp when determining if it needs to back
> up a file.
>
> My networker version is 6.1.2 (rebadged as solstice backup via
> SUN).
>
> Not that this is based on my research so far, I still need to
> do more testing but I hope someone can at least enlighten
> me as to the algorithm networker uses to determine when to
> backup a file when doing incremental backups.
>
> Regards,
> pdg
>
> --
>
> See mail headers for contact information.
>
> --
> Note: To sign off this list, send a "signoff" command via email
> to listserv AT listmail.temple DOT edu or visit the list's Web site at
> http://listmail.temple.edu/archives/networker.html where you can
> also view and post messages to the list.
> =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

<Prev in Thread] Current Thread [Next in Thread>