On (23/02/16 10:55), Harald Dunkel wrote:
>On 02/23/2016 10:00 AM, Jakub Hrozek wrote:
>> Typically, this happens when the machine SSSD is running on is very
>> busy, the sssd_be process is blocked writing some large result from
>> LDAP, the monitor process considers it stuck and kills it. However, we
>> /should/ restart and reconnect the subprocesses.
>Shoot the slow horse? Sorry to say, but I doubt that this is
>a reasonable approach. Can I turn off this feature?
>I would like to avoid to move my mailhost back to NIS client,
>but ypbind was never shot. Incoming EMails have been lost.
>What would you suggest?
I would rather focus on different thing.
Why is sssd_be process blocked for long time?
Do you use enumeration?
If yes do you really need it.
Workaround might be to increate timeout between heartbeats
which are used to ensure that the process is alive and capable of answering
requests. The default value is 10 seconds. Double it should be enough
because there is by default 6 heartbeats IIRC.
Manage your subscription for the Freeipa-users mailing list:
Go to http://freeipa.org for more info on the project