Interesting... that might well explain another "weird" problem I was seeing.
Time to add a few "-b" flags !
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On Friday, August 13th, 2021 at 5:56 PM, David Vasek <david.vasek(a)nic.cz> wrote:
That's because in normal (that is non-blocking)
mode knotc just
schedules operations in the server and reports whether the scheduling
was successful or not. You can start putting other operations in the
queue just after the command prompt returns. On the other hand, if you
want to wait until the operation finishes and immediately see a brief
result of it (or get an exit status reflecting the actual operation),
you must use the blocking mode (knotc -b zone-backup ...). See the last
paragraphs in the Note section here:
https://www.knot-dns.cz/docs/3.1/html/man_knotc.html
Regards,
David
On 2021-08-13 17:49, Laura Smith wrote:
> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
>
> On Friday, August 13th, 2021 at 12:37 PM, David Vasek
>
> david.vasek(a)nic.cz wrote:
>
> > Hi Laura,
> >
> > the ownership of the backup directory, its permissions and the user
> >
> > (and
> >
> > group) under which knotd is running are important here.
> >
> > Regards,
> >
> > David
>
> Thanks David, that worked.
>
> As an aside, I have noticed that the CLI exits 0 with "OK" even though
>
> the backup is still running (and so I have found problems calling tar
>
> immediately because "file being changed")