Author: markt
Date: Thu Aug 23 21:53:35 2018
New Revision: 1838767

URL: http://svn.apache.org/viewvc?rev=1838767&view=rev
Log:
Having read bz 62408 and performed some testing, update the docs to make the 
behaviour clear. This also re-enforces the need to apply the patch provided for 
bz 62408 which I hope to do shortly.

Modified:
    tomcat/jk/trunk/xdocs/miscellaneous/changelog.xml
    tomcat/jk/trunk/xdocs/reference/workers.xml

Modified: tomcat/jk/trunk/xdocs/miscellaneous/changelog.xml
URL: 
http://svn.apache.org/viewvc/tomcat/jk/trunk/xdocs/miscellaneous/changelog.xml?rev=1838767&r1=1838766&r2=1838767&view=diff
==============================================================================
--- tomcat/jk/trunk/xdocs/miscellaneous/changelog.xml (original)
+++ tomcat/jk/trunk/xdocs/miscellaneous/changelog.xml Thu Aug 23 21:53:35 2018
@@ -69,6 +69,10 @@
         of the current context path, it is impossible to implement this check
         without valid requests being rejected. (markt)
       </fix>
+      <add>
+        Clarify the behvaiour of lb workers when all ajp13 workers fail with
+        particular reference to the role of the retries attribute. (markt)
+      </add>
    </changelog>
   </subsection>
 </section>

Modified: tomcat/jk/trunk/xdocs/reference/workers.xml
URL: 
http://svn.apache.org/viewvc/tomcat/jk/trunk/xdocs/reference/workers.xml?rev=1838767&r1=1838766&r2=1838767&view=diff
==============================================================================
--- tomcat/jk/trunk/xdocs/reference/workers.xml (original)
+++ tomcat/jk/trunk/xdocs/reference/workers.xml Thu Aug 23 21:53:35 2018
@@ -628,9 +628,16 @@ This feature has been added in <b>jk 1.2
 <warn>This directive also exists for normal workers.
 For those it has a <a href="#Advanced Worker Directives">different 
meaning</a>.</warn>
 If the load balancer can not get a valid member worker or in case of failover,
-it will try again a number of times given by <b>retries</b>.
+it will try each member in turn a number of times given by <b>retries</b>.
 Before each retry, it will make a pause define by <b>retry_interval</b> 
directive.
 <p>
+Note that this means that, if all workers fail, there will be a total of
+number-of-workers * lb.retries * worker.retries requests before a 503 response
+is returned to the client. So for an lb worker with four members and a default
+configuration, if all workers fail there will be a total of 16 requests before
+a 503 response is returned to the client.
+</p>
+<p>
 Until version <b>1.2.16</b> the default value was 3.
 </p>
 </directive>



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

Reply via email to