[#173,!154] Added a note suggesting to set reconnect for all used backends.

diff --git a/doc/guide/dhcp4-srv.xml b/doc/guide/dhcp4-srv.xml
index 6c08867d8dc3ab093a17bf9abb254211af85d10a..7befc88735fb55d510b5c3f4b02339878c7372eb 100644 (file)
--- a/doc/guide/dhcp4-srv.xml
+++ b/doc/guide/dhcp4-srv.xml
@@ -505,6 +505,19 @@ The default value for MySQL and Postgres is 0, which disables automatic recovery
 causes the server to exit immediately upon detecting the loss of connectivity.
 The default value for Cassandra is 2000 ms.
     </para>
+
+    <note>
+      <para>
+        Automatic reconnect to database backends is configured individually per backend.
+        This allows you to tailor the recovery parameters to the each backend you use.
+        We do suggest that you either enable it for all backends or no backends so you
+        have consistent behavior.  Losing connectivity to a backend for which reconnect
+        is disabled will result in the server shutting itself down. This includes the
+        cases when lease database backend and hosts database backend is connected to
+        the same database instance.
+      </para>
+    </note>
+
   <para>Finally, the credentials of the account under which the server will
   access the database should be set:
 <screen>
@@ -673,17 +686,17 @@ causes the server to exit immediately upon detecting the loss of connectivity.
 The default value for Cassandra is 2000 ms.
     </para>
 
-    <warning>
+    <note>
       <para>
-        Current Kea release has a flaw whereby the values of <command>reconnect-wait-time</command>
-        and <command>max-reconnect-tries</command> specified in the <command>lease-database</command>
-        override the setting of those parameters in the <command>hosts-database</command>. The
-        parameters specified within the <command>hosts-database</command> are only effective
-        when <command>memfile</command> is used as a lease database backend. This behavior will
-        be corrected in the future releases of Kea.
+        Automatic reconnect to database backends is configured individually per backend.
+        This allows you to tailor the recovery parameters to the each backend you use.
+        We do suggest that you either enable it for all backends or no backends so you
+        have consistent behavior.  Losing connectivity to a backend for which reconnect
+        is disabled will result in the server shutting itself down. This includes the
+        cases when lease database backend and hosts database backend is connected to
+        the same database instance.
       </para>
-    </warning>
-
+    </note>
   </para>
   <para>Finally, the credentials of the account under which the server will
   access the database should be set:

diff --git a/doc/guide/dhcp6-srv.xml b/doc/guide/dhcp6-srv.xml
index 63be6fbd9dcd9c07a60b233bc808da008c90c6ff..968eee6006516d2f47b1952f74fd3fc762d574ad 100644 (file)
--- a/doc/guide/dhcp6-srv.xml
+++ b/doc/guide/dhcp6-srv.xml
@@ -502,6 +502,19 @@ The default value for MySQL and Postgres is 0, which disables automatic recovery
 causes the server to exit immediately upon detecting the loss of connectivity.
 The default value for Cassandra is 2000 ms.
   </para>
+
+  <note>
+    <para>
+      Automatic reconnect to database backends is configured individually per backend.
+      This allows you to tailor the recovery parameters to the each backend you use.
+      We do suggest that you either enable it for all backends or no backends so you
+      have consistent behavior.  Losing connectivity to a backend for which reconnect
+      is disabled will result in the server shutting itself down. This includes the
+      cases when lease database backend and hosts database backend is connected to
+      the same database instance.
+    </para>
+  </note>
+
   <para>
     Note that host parameter is used by MySQL and PostgreSQL
     backends. Cassandra has a concept of contact points that could be
@@ -610,16 +623,17 @@ causes the server to exit immediately upon detecting the loss of connectivity.
 The default value for Cassandra is 2000 ms.
   </para>
 
-  <warning>
-    <para>
-      Current Kea release has a flaw whereby the values of <command>reconnect-wait-time</command>
-      and <command>max-reconnect-tries</command> specified in the <command>lease-database</command>
-      override the setting of those parameters in the <command>hosts-database</command>. The
-      parameters specified within the <command>hosts-database</command> are only effective
-      when <command>memfile</command> is used as a lease database backend. This behavior will
-      be corrected in the future releases of Kea.
-    </para>
-  </warning>
+    <note>
+      <para>
+        Automatic reconnect to database backends is configured individually per backend.
+        This allows you to tailor the recovery parameters to the each backend you use.
+        We do suggest that you either enable it for all backends or no backends so you
+        have consistent behavior.  Losing connectivity to a backend for which reconnect
+        is disabled will result in the server shutting itself down. This includes the
+        cases when lease database backend and hosts database backend is connected to
+        the same database instance.
+      </para>
+    </note>
 
   <para>Finally, the credentials of the account under which the server will
   access the database should be set: