Date: Wed, 16 Jul 2008 13:01:46 +0530 From: "Tapan Chaudhari" <tapan.list@gmail.com> To: freebsd-hackers@freebsd.org Subject: Re: Can I change the device of the "/" mount point at boot time. Message-ID: <482257ad0807160031k34980a19na9895f38f125d4e0@mail.gmail.com> In-Reply-To: <200807151538.38285.doconnor@gsoft.com.au> References: <482257ad0807141106m679ec19frd853339637d27a2d@mail.gmail.com> <200807151302.30415.doconnor@gsoft.com.au> <482257ad0807142138j1f7b7dd8nda7de865a4616fc0@mail.gmail.com> <200807151538.38285.doconnor@gsoft.com.au>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Jul 15, 2008 at 11:38 AM, Daniel O'Connor <doconnor@gsoft.com.au> wrote: > On Tue, 15 Jul 2008, Tapan Chaudhari wrote: > > Thanks Deniel for the reply. I am aware of the fact you mentioned > > and will keep in mind. > > Well what i am trying to achieve is a simple thing to write an > > interception driver to catch all the i/os going to a particular > > device, do some manipulations on it and than let it through to the > > original device. Well as you mentioned about geom, I have recently > > posted a mail on GEOM mailing list as I could not find geom doing > > interception, the discussion is still on (You can see the mails with > > subject line "Can GEOM be used to intercept the I/o calls to an > > existing mounted device?"). Any sugessuions on interception driver > > will be helpful? > > My first question would be "Why do you want to do that?" I am planning to write a block level snapshot driver. > > > > As an interception driver is not possible, for time being I am going > > towards the redirection concept which will require a reboot and > > changing the devices on the mount points. For redirection driver, I > > dont think I will need geom. I can directly create a new device. > > Rather I think it would be an overhead using geom for a virtual > > device. > > Any thoughts on both the issues? > > I think you'd have a lower overhead and much less hassle writing a GEOM > class and using that. > This sounds good. I will try using GEOM first. But if I could achieve interception, as I described earlier, I will go for that rather than redirection. > > -- > Daniel O'Connor software and network engineer > for Genesis Software - http://www.gsoft.com.au > "The nice thing about standards is that there > are so many of them to choose from." > -- Andrew Tanenbaum > GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?482257ad0807160031k34980a19na9895f38f125d4e0>