v6 header format and differential backups

[XBrit]

Hey gurus.

Newsbin v5 is not friendly to differential disk backups. My SPOOL_V1 directory is about 35GB, which is probably far from the largest out there. Every time I read the latest headers in a.b.erotica and a couple other large-ish groups, it shuffles (just renames?) every directory for the groups in SPOOL_V1. Then my next differential backup backs up all 35GB, even though I think hardly any of it has really changed.

This kind of wastes disk space on my NAS box. Will v6 improve this at all?

[itimpi]

No - if anything it is likely to make it worse.

I would suggest you exclude the spool folder from any backups.

[XBrit]

No - if anything it is likely to make it worse.

I would suggest you exclude the spool folder from any backups.

LOL, but I don't want to lose that spool data if my RAID-0 crashes! Took a long time to download all that stuff.

Anyway, whatever. I can live with it. Just asking.

[Quade]

it shuffles (just renames?) every directory for the groups in SPOOL_V1. Then my next differential backup backs up all 35GB, even though I think hardly any of it has really changed.

I don't get what you're talking about here. The names don't change in the spool_v1 folder. Maybe you mean timestamps.

[XBrit]

it shuffles (just renames?) every directory for the groups in SPOOL_V1. Then my next differential backup backs up all 35GB, even though I think hardly any of it has really changed.

I don't get what you're talking about here. The names don't change in the spool_v1 folder. Maybe you mean timestamps.

Yes, very possibly. That would also be enough to provoke a full backup when differential was requested.

[DThor]

Just as a recommend, if you're at the point where you're concerned over the volume of spools, which I can certainly see for many, then I would definitely exclude them. Constantly backing up many gigs of data that will always exist out there to cover the unlikely event of a drive crash, well, that's a lot of bit moving for not a lot of benefit. Especially when the solution is as simple as hitting a button plus time, it's not like it's some unique data that is difficult to reproduce. Just sayin'...

DT