[OpenAFS] vos shadow to backup user homes

Mon, 26 Aug 2013 07:20:38 -0700

--047d7b67643ae3dcba04e4da7514
Content-Type: text/plain; charset=ISO-8859-1

I usually mount not only /vicepxx but also /usr/afs & /usr/vice on the raid
drives, that way a motherboard upgrade requires swapping the controller &
drives.

A consistent problem is raid driver upgrades which for the low end
controllers never happen.

On Mon, Aug 26, 2013 at 6:46 AM, <stephen@physics.unc.edu> wrote:

> Hi Shouri,
>
> I've used shadow volumes in the past. Thankfully I never had to test
> bringing an entire shadow server into production. I also plan to use it in
> the near future (barring contrary advice from the list) for a server which
> houses dozens of TBs of research data for a project which can suffer some
> downtime (as long as it's unavoidable) but not as much downtime as would be
> needed to restore dozens of ~1TB volumes.
>
> I'll leave the topic open though, and would welcome comments on shadow
> volumes from any of the devs.
>
> Regarding rsync, depending on the size of your partitions, its performance
> may make you cry. I use an rsync-based backup application for my non-AFS
> data and am starting to surpass the limits of what rsync can do in a
> reasonable amount of time and RAM on current hardware.
>
> However the main reason I'm replying is your comment about RAID. IMO,
> anytime you're configuring a mission-critical system without RAID you're
> probably asking for future headaches. I think the only time I'd consider it
> is if the system had no unique data on it and could be made part of a HA
> cluster using [heartbeat, etc]. But at that point you're just abstracting
> your redundancy at a different level.
>
> All of my database and fileservers currently use hardware raid (3ware or
> LSI/PERC). But one of my "idle time" projects -- a bit of an inside joke
> since I have no idle time -- is to play around with ZFS on linux to see if
> I feel it's ready for prime time yet or not.
>
> PS. I noticed "Associate Professor" in your signature. Have you consulted
> with your local IT support? If they're not overworked, they may have
> additional advice specific to your site, discipline, etc.
>
> Cheers,
> Stephen
>
> On Sun, 25 Aug 2013, Shouri Chatterjee wrote:
>
>
>> Dear All,
>>
>> I wanted to ask about "vos shadow" and whether it is being used as a
>> solution on production systems to back-up user home directories.
>>
>> The most significant information I can find is a thread from this email
>> archive last year. http://lists.openafs.org/**
>> pipermail/openafs-info/2012-**December/039077.html<http://lists.openafs.org/pipermail/openafs-info/2012-December/039077.html>
>>
>> I can dedicate a server to host only shadow volumes. If an active server
>> fails and dies, the shadow copy can be brought online. Is this a better
>> solution than, say:
>> (1) keeping a periodically rsync'd copy of the /vicepx partitions on a
>> shadow server
>> (2) afs over drbd
>>
>> I am trying to use commodity hardware (no RAID, no scope for software
>> RAID either) to build cheap AFS storage.
>>
>> Shouri
>>
>> ____________
>>
>> Shouri Chatterjee
>> Associate Professor
>> Department of Electrical Engineering
>> IIT Delhi, Hauz Khas
>> New Delhi 110016
>> India
>>
>> Phone:  +91 11 2659 1099 (O)
>>         +91 11 2659 1619 (R)
>>
>>
>> ______________________________**_________________
>> OpenAFS-info mailing list
>> OpenAFS-info@openafs.org
>> https://lists.openafs.org/**mailman/listinfo/openafs-info<https://lists.openafs.org/mailman/listinfo/openafs-info>
>>
>>  ______________________________**_________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/**mailman/listinfo/openafs-info<https://lists.openafs.org/mailman/listinfo/openafs-info>
>

--047d7b67643ae3dcba04e4da7514
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>I usually mount not only /vicepxx but also /usr/afs &=
amp; /usr/vice on the raid drives, that way a motherboard upgrade requires =
swapping the controller &amp; drives. </div>A consistent problem is =
raid driver upgrades which for the low end controllers never happen. 
</div><div class=3D"gmail_extra"> <div class=3D"gmail_quote">On Mon,=
 Aug 26, 2013 at 6:46 AM, &lt;<a href=3D"mailto:stephen@=
physics.unc.edu" target=3D"_blank">stephen@physics.unc.edu</a>&gt; w=
rote: 
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">Hi Shouri, 
 
I&#39;ve used shadow volumes in the past. Thankfully I never had to test br=
inging an entire shadow server into production. I also plan to use it in th=
e near future (barring contrary advice from the list) for a server which ho=
uses dozens of TBs of research data for a project which can suffer some dow=
ntime (as long as it&#39;s unavoidable) but not as much downtime as would b=
e needed to restore dozens of ~1TB volumes.

I&#39;ll leave the topic open though, and would welcome comments on shadow =
volumes from any of the devs. 
 
Regarding rsync, depending on the size of your partitions, its performance =
may make you cry. I use an rsync-based backup application for my non-AFS da=
ta and am starting to surpass the limits of what rsync can do in a reasonab=
le amount of time and RAM on current hardware.

However the main reason I&#39;m replying is your comment about RAID. IMO, a=
nytime you&#39;re configuring a mission-critical system without RAID you&#3=
9;re probably asking for future headaches. I think the only time I&#39;d co=
nsider it is if the system had no unique data on it and could be made part =
of a HA cluster using [heartbeat, etc]. But at that point you&#39;re just a=
bstracting your redundancy at a different level.

All of my database and fileservers currently use hardware raid (3ware or LS=
I/PERC). But one of my &quot;idle time&quot; projects -- a bit of an inside=
 joke since I have no idle time -- is to play around with ZFS on linux to s=
ee if I feel it&#39;s ready for prime time yet or not.

PS. I noticed &quot;Associate Professor&quot; in your signature. Have you c=
onsulted with your local IT support? If they&#39;re not overworked, they ma=
y have additional advice specific to your site, discipline, etc. 
 
Cheers, 
Stephen 
 
On Sun, 25 Aug 2013, Shouri Chatterjee wrote: 
 
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
 
Dear All, 
 
I wanted to ask about &quot;vos shadow&quot; and whether it is being used a=
s a solution on production systems to back-up user home directories. 
 
The most significant information I can find is a thread from this email arc=
hive last year. <a href=3D"http://lists.openafs.org/pipermail/openafs-info/=
2012-December/039077.html" target=3D"_blank">http://lists.openafs.org/</=
u>pipermail/openafs-info/2012-December/039077.html</a>

I can dedicate a server to host only shadow volumes. If an active server fa=
ils and dies, the shadow copy can be brought online. Is this a better solut=
ion than, say: 
(1) keeping a periodically rsync&#39;d copy of the /vicepx partitions on a =
shadow server 
(2) afs over drbd 
 
I am trying to use commodity hardware (no RAID, no scope for software RAID =
either) to build cheap AFS storage. 
 
Shouri 
 
____________ 
 
Shouri Chatterjee 
Associate Professor 
Department of Electrical Engineering 
IIT Delhi, Hauz Khas 
New Delhi 110016 
India 
 
Phone: =A0<a href=3D"tel:%2B91%2011%202659%201099" value=3D"+911126591099" =
target=3D"_blank">+91 11 2659 1099</a> (O) 
=A0 =A0 =A0 =A0 <a href=3D"tel:%2B91%2011%202659%201619" value=3D"+91112659=
1619" target=3D"_blank">+91 11 2659 1619</a> (R) 
 
 
_______________________________________________ 
OpenAFS-info mailing list 
<a href=3D"mailto:OpenAFS-info@openafs.org" target=3D"_blank">OpenAFS-info@=
openafs.org</a> 
<a href=3D"https://lists.openafs.org/mailman/listinfo/openafs-info" target=
=3D"_blank">https://lists.openafs.org/mailman/listinfo/openafs-info<=
/a> 
 
</blockquote>
_______________________________________________ 
OpenAFS-info mailing list 
<a href=3D"mailto:OpenAFS-info@openafs.org" target=3D"_blank">OpenAFS-info@=
openafs.org</a> 
<a href=3D"https://lists.openafs.org/mailman/listinfo/openafs-info" target=
=3D"_blank">https://lists.openafs.org/mailman/listinfo/openafs-info<=
/a> 
</blockquote></div> </div>

--047d7b67643ae3dcba04e4da7514--