[Box Backup] manual snapshot problem

[Chris Wilson]

Hi Matt and all,

On Sat, 25 Oct 2008, Matt Brown wrote:

> >I remember that Matt noticed that when not running automatic backups 
> >(i.e. in snapshot mode), when an error occurs during the backup which 
> >causes it not to complete successfully, it would not have retried 
> >automatically, and I think I fixed that behaviour.
> 
> Yes, the behavior was if a backup failed in Snapshot mode at any point 
> during the sync - the backup would cease and not restart. However since 
> you added the retry part from Lazy mode all is well and I can confirm it 
> does in fact retry flawlessly.

OK, good.

My question to all users is whether it should retry immediately (as it 
does now), or after some configurable delay (which would solve the issue 
that Torsten reported, and could be set to zero if you want an immediate 
retry).

> Maybe a RetryOnFailure (or similar) option for those who do not wish to 
> automatically retry in case of failure/error, however I would wonder if 
> anyone would want that option

Me too. I can't see why a user would disable that option, and I woulnd't 
like to introduce a new option if the default behaviour can be made more 
suitable instead.

Perhaps we should treat errors that require user intervention (such as 
non-network errors) differently and not retry in those cases. I haven't 
decided yet and would appreciate input.

> as this is (IIRC) the default behavior in Lazy mode ?

Yes, but this is all about snapshot mode. The configuration option would 
have no effect in Lazy Mode (or rather, UpdateStoreInterval would retain 
its usual function in lazy mode).

> but as for behavior of retries via UpdateStoreInterval or similar is 
> fine with me as long as the resultant is it retries the failed 
> session/sync :)

OK, great.

The effect of the change that I'm proposing is that we redefine Snapshot 
Mode:

* the original meaning was "start a single backup run now, and then go 
  idle whether it succeeds or fails".

* the current meaning is "start backing up and keep trying until you 
  finish successfully, then go idle"

* proposed new meaning 1: "start backing up and switch to lazy mode until 
  you finish a backup successfully, then go idle"

* proposed new meaning 2: "start backing up and switch to lazy mode until 
  you encounter an error that's not a network error, then go idle"

Does any of these proposals make particular sense or nonsense to our 
users?

Cheers, Chris.
-- 
_____ __     _
\  __/ / ,__(_)_  | Chris Wilson <0000 at qwirx.com> - Cambs UK |
/ (_/ ,\/ _/ /_ \ | Security/C/C++/Java/Ruby/Perl/SQL Developer |
\ _/_/_/_//_/___/ | Stop nuclear war http://www.nuclearrisk.org |