[AFS3-std] Snapshot Identifiers

[Derrick Brashear]

--001636163f41ec3a91046cf4820c
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

On Mon, Jun 22, 2009 at 4:29 AM, Hartmut Reuter <reuter@rzg.mpg.de> wrote:

>
> With these giant numbers of snapshots I see some other problems:
>
> If there should coexist multiple snapshots in the same partition
> belonging to the same volume group the present design of the linktable
> of the NAMEI-fileserver will break because it allows only a limited
> number of versions (5) and a limited value for the link count (7) per
> version of a vnode number's data.
>

that's not a protocol issue.


>
> Also if the volume id should become 64 bit certainly the NAMEI mapping
> algorithm of volume ids to directory trees would have to change.
> This change would not only affect the fileserver, volserver and
> salvager, but also rxosd.
>

same deal

--001636163f41ec3a91046cf4820c
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<br><br><div class=3D"gmail_quote">On Mon, Jun 22, 2009 at 4:29 AM, Hartmut=
 Reuter <span dir=3D"ltr">&lt;<a href=3D"mailto:reuter@rzg.mpg.de">reuter@r=
zg.mpg.de</a>&gt;</span> wrote:<br><blockquote class=3D"gmail_quote" style=
=3D"border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; p=
adding-left: 1ex;">
<br>
With these giant numbers of snapshots I see some other problems:<br>
<br>
If there should coexist multiple snapshots in the same partition<br>
belonging to the same volume group the present design of the linktable<br>
of the NAMEI-fileserver will break because it allows only a limited<br>
number of versions (5) and a limited value for the link count (7) per<br>
version of a vnode number&#39;s data.<br>
</blockquote><div><br>that&#39;s not a protocol issue.<br>=A0<br></div><blo=
ckquote class=3D"gmail_quote" style=3D"border-left: 1px solid rgb(204, 204,=
 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
Also if the volume id should become 64 bit certainly the NAMEI mapping<br>
algorithm of volume ids to directory trees would have to change.<br>
This change would not only affect the fileserver, volserver and<br>
salvager, but also rxosd.<br>
</blockquote><div><br>same deal <br></div></div><br>

--001636163f41ec3a91046cf4820c--