On 05/10/2016 04:00 PM, Jakub Hrozek wrote:

On Thu, Apr 21, 2016 at 02:54:21PM +0200, Pavel Březina wrote:

...

We can fail in sasl_bind_send() with ERR_AUTH_FAILED for basically unspecified reason but we do not failover to next server. This patch should fix it.

As said on the meeting, I didn't reproduce it and I'm not sure if it will fix the customer issue unless they confirm it, but it seems to be a valid patch anyway.

...

From c9e3bf2726d9edba3cddc91c4ae98a87a5c1652a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pavel=20B=C5=99ezina?= pbrezina@redhat.com Date: Wed, 6 Apr 2016 13:20:59 +0200 Subject: [PATCH 1/2] Inform about (un)successful connection

---

src/providers/ldap/sdap_async_connection.c | 5 +++++ 1 file changed, 5 insertions(+)

diff --git a/src/providers/ldap/sdap_async_connection.c b/src/providers/ldap/sdap_async_connection.c index f9074afb0c1340c7c2a50d4df0021eee4ae0d076..435cb1bb6864b516d386fafa360d062f780b5d04 100644 --- a/src/providers/ldap/sdap_async_connection.c +++ b/src/providers/ldap/sdap_async_connection.c @@ -2035,6 +2035,9 @@ int sdap_cli_connect_recv(struct tevent_req *req, if (tevent_req_is_error(req, &tstate, &err)) { /* mark the server as bad if connection failed */ if (state->srv) {

•        DEBUG(SSSDBG_OP_FAILURE, "Unable to establish connection "
  

•              "[%d]: %s\n", err, sss_strerror(err));
  

In file included from /sssd/src/util/util.h:54:0, from /sssd/src/providers/ldap/sdap_async_connection.c:26: /sssd/src/providers/ldap/sdap_async_connection.c: In function 'sdap_cli_connect_recv': /sssd/src/providers/ldap/sdap_async_connection.c:2038:38: warning: format '%d' expects argument of type 'int', but argument 6 has type 'uint64_t {aka long unsigned int}' [-Wformat=] DEBUG(SSSDBG_OP_FAILURE, "Unable to establish connection " ^ /sssd/src/util/debug.h:108:22: note: in definition of macro 'DEBUG' format, ##__VA_ARGS__); \

Fixed.

                   ^

...

          be_fo_set_port_status(state->be, state->service->name,
                                state->srv, PORT_NOT_WORKING);
      } else {

@@ -2048,6 +2051,8 @@ int sdap_cli_connect_recv(struct tevent_req *req, } return EIO; } else if (state->srv) {

•    DEBUG(SSSDBG_TRACE_FUNC, "Connection established.\n");
  

•     be_fo_set_port_status(state->be, state->service->name,
                            state->srv, PORT_WORKING);
  }
  

-- 2.1.0

...

From 8d6f024ac77ddb8dc87d7fd693cfe5846f3a88d6 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pavel=20B=C5=99ezina?= pbrezina@redhat.com Date: Wed, 6 Apr 2016 13:21:21 +0200 Subject: [PATCH 2/2] Failover to next server if authentication fails

I was only able to test this patch 'synthetically', by setting ret to the ERR_AUTH_FAILED value. But then it seemed to do the right thing, so ACK.

Here as well.

On Wed, May 11, 2016 at 11:42:40AM +0200, Jakub Hrozek wrote:

On Wed, May 11, 2016 at 10:24:10AM +0200, Pavel Březina wrote:

...

On 05/10/2016 04:00 PM, Jakub Hrozek wrote:

...

On Thu, Apr 21, 2016 at 02:54:21PM +0200, Pavel Březina wrote:

...

We can fail in sasl_bind_send() with ERR_AUTH_FAILED for basically unspecified reason but we do not failover to next server. This patch should fix it.

As said on the meeting, I didn't reproduce it and I'm not sure if it will fix the customer issue unless they confirm it, but it seems to be a valid patch anyway.

...

From c9e3bf2726d9edba3cddc91c4ae98a87a5c1652a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pavel=20B=C5=99ezina?= pbrezina@redhat.com Date: Wed, 6 Apr 2016 13:20:59 +0200 Subject: [PATCH 1/2] Inform about (un)successful connection

---

src/providers/ldap/sdap_async_connection.c | 5 +++++ 1 file changed, 5 insertions(+)

diff --git a/src/providers/ldap/sdap_async_connection.c b/src/providers/ldap/sdap_async_connection.c index f9074afb0c1340c7c2a50d4df0021eee4ae0d076..435cb1bb6864b516d386fafa360d062f780b5d04 100644 --- a/src/providers/ldap/sdap_async_connection.c +++ b/src/providers/ldap/sdap_async_connection.c @@ -2035,6 +2035,9 @@ int sdap_cli_connect_recv(struct tevent_req *req, if (tevent_req_is_error(req, &tstate, &err)) { /* mark the server as bad if connection failed */ if (state->srv) {

•        DEBUG(SSSDBG_OP_FAILURE, "Unable to establish connection "
  

•              "[%d]: %s\n", err, sss_strerror(err));
  

In file included from /sssd/src/util/util.h:54:0, from /sssd/src/providers/ldap/sdap_async_connection.c:26: /sssd/src/providers/ldap/sdap_async_connection.c: In function 'sdap_cli_connect_recv': /sssd/src/providers/ldap/sdap_async_connection.c:2038:38: warning: format '%d' expects argument of type 'int', but argument 6 has type 'uint64_t {aka long unsigned int}' [-Wformat=] DEBUG(SSSDBG_OP_FAILURE, "Unable to establish connection " ^ /sssd/src/util/debug.h:108:22: note: in definition of macro 'DEBUG' format, ##__VA_ARGS__); \

Fixed.

...

                   ^

...

          be_fo_set_port_status(state->be, state->service->name,
                                state->srv, PORT_NOT_WORKING);
      } else {

@@ -2048,6 +2051,8 @@ int sdap_cli_connect_recv(struct tevent_req *req, } return EIO; } else if (state->srv) {

•    DEBUG(SSSDBG_TRACE_FUNC, "Connection established.\n");
  

•     be_fo_set_port_status(state->be, state->service->name,
                            state->srv, PORT_WORKING);
  }
  

-- 2.1.0

...

From 8d6f024ac77ddb8dc87d7fd693cfe5846f3a88d6 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pavel=20B=C5=99ezina?= pbrezina@redhat.com Date: Wed, 6 Apr 2016 13:21:21 +0200 Subject: [PATCH 2/2] Failover to next server if authentication fails

I was only able to test this patch 'synthetically', by setting ret to the ERR_AUTH_FAILED value. But then it seemed to do the right thing, so ACK.

Here as well.

ACK.

CI is still running. I filed https://fedorahosted.org/sssd/ticket/3010 to track this issue..

CI: http://sssd-ci.duckdns.org/logs/job/43/18/summary.html

master: 8eb4a1f773b6c2246b0a6c6e3c3b8ef4d79c9ca6 73dd89c3fb361dab43b4802510f4c64d282dbde1