Date: Mon, 15 Jan 2001 10:02:02 +0100 (MET) From: Jean-Francois Dockes <jean-francois.dockes@wanadoo.fr> To: mjacob@feral.com Cc: Eric Lee Green <eric@estinc.com>, freebsd-scsi@FreeBSD.ORG Subject: Re: Why filemarks in sardpos? Message-ID: <14946.48266.332620.887204@hautmedoc.dockes.com> In-Reply-To: <Pine.BSF.4.21.0101132006280.14728-100000@beppo.feral.com> References: <Pine.LNX.4.21.0101131547160.6204-100000@h23.estsatnet> <Pine.BSF.4.21.0101132006280.14728-100000@beppo.feral.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Matthew Jacob writes: > [cut] > So, all other things being equal, "First Block Value" should be > good enough if one assumes that blocks Z..A get flushed to tape. > > ... > Anyone else out there have an opinion on this? > ... Funny, I was just working with sa last week for the purpose of porting some NDMP backup tape server code. I had trouble with how 'sa' maintains the current tape position (I submitted a few PRs), and I wondered if I could use MTIOCRDSPOS instead, but I was prevented to do this precisely by the flush thing (I didn't even try to use MTIOCRDSPOS, as the code made it clear that it was unusable for my purpose). So, yes, I have an opinion, and it is that no flush should be performed. In practise, I think that only the 'First Block' value is useful for backup software, and its signification is pretty clear. I guess that the comment in the code about how the SCSI spec is vague relate to the 'Last block' value, but I can't really understand what the 'Last Block' thing is good for anyway. Jean-Francois Dockes To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14946.48266.332620.887204>