On Tue, Jun 02, 2015 at 10:56:03AM +0200, Lukas Slebodnik wrote:
>> >+static void test_get_next_domain_disabled(void
**state)
>> >+{
>> >+ struct dom_list_test_ctx *test_ctx = talloc_get_type(*state,
>> >+ struct
dom_list_test_ctx);
>> >+ struct sss_domain_info *dom = NULL;
>> >+
>> >+ for (dom = test_ctx->dom_list; dom; dom = get_next_domain(dom,
true)) {
>> >+ dom->disabled = true;
>> >+ }
>> it would be good to have more test with combination of some disabled domains
>> and with or without subdomain.
>>
>> There I have a question.
>> If main domain is disabled should we return its enabled subdomains?
>> Does it make sense?
>
>I'm not sure. So far the main domain cannot be disabled, only offline.
>Only subdomains can be disabled when processing newly updated subdomain
>list. I also think the disabled flag should only be readable by the
>domain utility module and maybe the subdomains code, not the deamon.
>
>I think this question will be more valid once we start supporting drop-in
>configurations. Then, during processing I guess we would disable even main
>existing domains and start enabling the new ones.
>
Fair enough
http://sssd-ci.duckdns.org/logs/job/16/61/summary.html
ACK
LS
* master: 4f97aaa2fd81e90c2d1f74d8b6619083db4061a5