On Fri, Sep 15, 2017 at 5:54 PM, Patrick Uiterwijk
<puiterwijk(a)redhat.com> wrote:
Hi all,
After a lot of debugging, it seems IPA 4.5.0 broke active/active
failover for krb5.
While I wait on getting that fixed, I would like to request +1s for
the following patch.
commit 4005fd5929c034436e21c56af0322d53cef04e74
Author: Patrick Uiterwijk <puiterwijk(a)redhat.com>
Date: Fri Sep 15 22:50:02 2017 +0000
Fix krb5 with failover
Seems like IPA 4.5.0 broke active/active failover of krb5 KDC.
While we wait on getting that fixed, let's set us up for
active/passive failover on the HTTPD end.
Since we can't do active/passive for UDP (there's no checks
there), let's just remove ipa02 for those.
Signed-off-by: Patrick Uiterwijk <puiterwijk(a)redhat.com>
diff --git a/roles/haproxy/templates/haproxy.cfg
b/roles/haproxy/templates/haproxy.cfg
index be1e5b5..cda10ab 100644
--- a/roles/haproxy/templates/haproxy.cfg
+++ b/roles/haproxy/templates/haproxy.cfg
@@ -340,7 +340,7 @@ listen ipa 0.0.0.0:10053
balance hdr(appserver)
server ipa01 ipa01:443 check inter 10s rise 1 fall 2 ssl verify
required ca-file /etc/haproxy/ipa.pem
{% if env != "staging" %}
- server ipa02 ipa02:443 check inter 10s rise 1 fall 2 ssl verify
required ca-file /etc/haproxy/ipa.pem
+ server ipa02 ipa02:443 check inter 10s rise 1 fall 2 ssl verify
required ca-file /etc/haproxy/ipa.pem backup
{% endif %}
option httpchk GET /ipa/ui/
@@ -354,7 +354,7 @@ listen krb5 0.0.0.0:1088
timeout connect 86400000
server ipa01 ipa01:88 weight 1 maxconn 16384
{% if env == "production" %}
- server ipa02 ipa02:88 weight 1 maxconn 16384
+ # server ipa02 ipa02:88 weight 1 maxconn 16384
{% endif %}
listen docker-candidate-registry 0.0.0.0:10054
+1
-AdamM
> _______________________________________________
> infrastructure mailing list -- infrastructure(a)lists.fedoraproject.org
> To unsubscribe send an email to infrastructure-leave(a)lists.fedoraproject.org