[OpenAFS] Re: backup strategy
Kristen J. Webb
Tue, 25 Nov 2014 13:19:25 -0700
I want to thank Andrew and Garrett for their honorable mention of our TiBS
backup solution. Here is some additional information relevant to the discussion
they have brought up.
TiBS easily supports large AFS volumes. There are many factors that can limit
quota sizes such as the size of the vice partitions, balancing constraints, and
of course the impact that backup can have. The deployment of TiBS to back up
AFS has generally resulted in customers being able to increase quota sizes. This
is primarily due to the removal of large, periodic full backups from the cell
using our synthetic backup capability. If you deploy your file servers with
large vice partitions (tens of terabytes each), then 2TB volume sizes shouldn't
be a problem.
We currently support incremental restore for an individual file. Our online
media database can be searched to determine exactly which incremental backup
volume needs to be restored to obtain the specific version of a file. When
restoring a sub-directory, a full volume restore is still required.
When considering cost, our software has technologies that can reduce costs in
other areas that are worth noting:
1. True incremental backups only take changes from a volume since the last
time it was updated and backed up. Backup windows are short, and cell resources
are returned to the actual users. The vos dump interface is used to collect
data through the volserver, keeping the backup traffic out of the fileserver
process that regular users access.
2. The Full Version of our software supports multiple level synthetic
backups that reduce backup server workload and backup media storage costs.
Adding a backup level to generate full backups every 6 months instead of every
month can have a huge impact on how much backup hardware and storage media is
3. Size based scheduling (versus traditional time based approaches) has two
major advantages. Periodic synthetic full backups of unchanged or little
changed data can be deferred indefinitely reducing backup server loads and long
term backup media storage costs. Very active volumes are identified and
consolidated more quickly, further reducing backup server loads and costs for
short term media storage.
There are many other advantages to performing synthetic backups that our
customers enjoy. I'd like to mention a few that I believe are among the most
- 80-90% or more of backup operations occur on dedicated backup servers
with zero impact on the rest of the network and computing environment, outside
of desired backup windows. Basically, our backup servers can run all day.
- Synthetic backup is like an on-going restore. It checks that your backup
media and storage devices are working properly and can detect issue with disk,
tape devices and tape media in advance of a real restore request.
- Synthetic backups can be repaired. You simply cannot do this with a
system that only performs network based vos dumps.
I also took a look at the tsmafs module and had a couple technical questions
about how it works.
1. What happens when you rename a directory (either a volume name or within
a volume)? Does the entire directory need to get backed up again?
2. While the use of hard links is limited in AFS, will tsmafs preserve
those links on restore?
On 11/18/14, 2:12 PM, Garrett Wollman wrote:
> <<On Mon, 10 Nov 2014 18:19:31 -0600, Andrew Deason <email@example.com> said:
>> Teradactyl's TiBS:
>> IIRC Teradactyl likes to trumpet their "synthetic" full dumps, which is
>> a feature where they use existing incremental and full dumps to generate
>> a new "full" dump every so often. This addresses what you were talking
>> about before, because it avoids needing to retain e.g. hundreds of daily
>> incrementals, but also avoids needing to periodically dump all data at
> We use TiBS and are quite happy with it. It understands the "vos
> dump" format and is able to index dumps and make synthetic backups as
> you describe. However, it is not able to do file-by-file restores --
> the entire volume must be restored. For this reason we have a policy
> restriction on the size of AFS volumes, to ensure that we always have
> enough free space to do restores.
> On the down side, it is quite expensive (as most enterprise backup
> systems are) -- but we get good service from Teradactyl, and they
> often report problems to us before we know about them. However, we
> probably wouldn't be using TiBS if we did not also have half a
> petabyte of NFS storage and servers to back up.
> OpenAFS-info mailing list
This message is NOT encrypted
Mr. Kristen J. Webb
Chief Technology Officer
2450 Baylor Dr. S.E.
Albuquerque, New Mexico 87106
Providers of Scalable Backup Solutions
for Unique Data Environments
NOTICE TO RECIPIENTS: Any information contained in or attached to this message
is intended solely for the use of the intended recipient(s). If you are not the
intended recipient of this transmittal, you are hereby notified that you
received this transmittal in error, and we request that you please delete and
destroy all copies and attachments in your possession, notify the sender that
you have received this communication in error, and note that any review or
dissemination of, or the taking of any action in reliance on, this communication
is expressly prohibited.
Regular internet e-mail transmission cannot be guaranteed to be secure or
error-free. Therefore, we do not represent that this information is complete or
accurate, and it should not be relied upon as such. If you prefer to communicate
with Teradactyl LLC. using secure (i.e., encrypted and/or digitally signed)
e-mail transmission, please notify the sender. Otherwise, you will be deemed to
have consented to communicate with Teradactyl via regular internet e-mail
transmission. Please note that Teradactyl reserves the right to intercept,
monitor, and retain all e-mail messages (including secure e-mail messages) sent
to or from its systems as permitted by applicable law.