Chris chris2014 at postbox.xyz
Mon Dec 24 09:56:37 UTC 2018

On Mon, 24 Dec 2018 02:05:19 +0100
Peter Eriksson via samba wrote:

> I’m not sure if it will help you but we’re seeing something similar
> that I call the “10 hours problem” here. On any Samba - after 4.7.6
> (we’re currently trying out 4.9.4 but still no go) - at 10 hours
> after it (smbd & winbindd) was restarted it new connections often
> (but not 100% all the time, but we think it is related to the number
> of connections, or active connections, or how many are encrypted or
> something, since we’re not seeing it on the test servers with zero or
> just one or two connections).

Thank you for your reply!

I'm not sure, if we're having the same issue, because:

- using rid instead of ad backend
- we've only 60 users and one group
- no nested groups of course
- timeout is always service ticket period. If it is set to 10 min.,
  timeout is after 10 min.
- there seems to be no timeout with NTLM auth, but haven't tested this
  for > 10 h
- I don't restart smbd
- it seems only version 4.5 from debian stretch is affected, not 4.8
  from Louis van Belle
- interruption is 100% all the time when kerberos is used
- doesn't matter how many connections are open. Only one connection on
  my test machine is sufficient.

Thank you for your script. Maybe I'll do some further tests between the

- Chris

More information about the samba mailing list