Re: [gentoo-user] [OT] SMR drives (WAS: cryptsetup close and device in use when it is not)

[William Kenworthy <billk@iinet.net.au>]

On 3/8/21 5:52 am, Frank Steinmetzger wrote:
> Am Mon, Aug 02, 2021 at 01:38:31PM +0800 schrieb William Kenworthy:
>
>>> Yup. Today I did my (not so) weekly backup and rsynced the repo to the new
>>> drive. After that I wanted to compare performance of my old 3 TB drive and
>>> the new SMR one by deleting a snapshot from the repo on each drive. But Borg
>>> objected on the second deletion, because “the cache was newer”. But that’s
>>> okay. I actually like this, as this will prevent me from chaning two repos
>>> in parallel which would make them incompatible.
>>>
>> Keep in  mind that both repos have the same ID - you should also rsync
>> the cache and security directories as well as they are now out of sync
>> (hence the warning).
> That thought crossed my mind recently but I was unsure how to store the
> cache. But since the repo is a monolith, it should suffice to rsync
> the whole cache directory to the backup drive (or do it as a tar).
>
> The only problem is the temporal sequence:
> 1. Host A runs borg and gets a current cache.
> 2. Host B runs borg on the same repo and gets a current cache.
>   2a. Host A now has an outdated cache.
>
> Usually, Host B uses Host A via ssh as remote location of the repository.
> So I could simply run a borg command on Host A to update the cache somehow.
>
>> Be very careful on how you do this - you are one step away from losing the
>> while repo if the cache gets out of sync.  The docs warn against rsyncing
>> two repos and then using them at the same time for a good reason.
> I won’t use them at the same time. It will always be one direction:
> Hosts --[borg]--> Main backup drive --[rsync]--> secondary backup drive
>
You could delete and rebuild the cache each time (or I think there is a
way to do without it).  There are quite a few threads on the borg lists
about this in the past (usually people trying to recover trashed repos)
- you might ask there if there is a way to deal with changing the ID now?

In any case, I think doing it the way you are has a fairly high chance
you will irretrievably trash both repos.

BillK