Author: mturk
Date: Fri Sep 1 12:15:18 2006
New Revision: 439427
URL: http://svn.apache.org/viewvc?rev=439427&view=rev
Log:
Added force_recovery for lb worker.
If all workers of the loadbalancer are in error state,
force the recovery on first attempt disregarding
maintain time. This fixes the gap where lb was
returning 503 and the Tomcat was up, but the recovery
was not yet scheduled.
Modified:
tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c
Modified: tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c
URL:
http://svn.apache.org/viewvc/tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c?rev=439427&r1=439426&r2=439427&view=diff
==============================================================================
--- tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c (original)
+++ tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c Fri Sep 1 12:15:18
2006
@@ -258,6 +258,30 @@
JK_TRACE_EXIT(l);
}
+static int force_recovery(lb_worker_t *p,
+ jk_logger_t *l)
+{
+ unsigned int i;
+ int forced = 0;
+ worker_record_t *w = NULL;
+ JK_TRACE_ENTER(l);
+
+ for (i = 0; i < p->num_of_workers; i++) {
+ w = &p->lb_workers[i];
+ if (w->s->state == JK_LB_STATE_ERROR) {
+ if (JK_IS_DEBUG_LEVEL(l))
+ jk_log(l, JK_LOG_INFO,
+ "worker %s is marked for recovery",
+ w->s->name);
+ w->s->state = JK_LB_STATE_RECOVER;
+ forced++;
+ }
+ }
+
+ JK_TRACE_EXIT(l);
+ return forced;
+}
+
/* Divide old load values by the decay factor,
* such that older values get less important
* for the routing decisions.
@@ -290,6 +314,7 @@
JK_TRACE_ENTER(l);
if (p && p->worker_private) {
+ unsigned int n = 0;
lb_worker_t *lb = (lb_worker_t *)p->worker_private;
for (i = 0; i < lb->num_of_workers; i++) {
@@ -318,6 +343,14 @@
recover_workers(lb, curmax, now, l);
}
+ for (i = 0; i < lb->num_of_workers; i++) {
+ if (lb->lb_workers[i].s->state != JK_LB_STATE_ERROR) {
+ ++n;
+ }
+ }
+ if (!n) {
+ force_recovery(lb, l);
+ }
jk_shm_unlock();
}
@@ -857,10 +890,35 @@
}
else {
/* NULL record, no more workers left ... */
- jk_log(l, JK_LOG_ERROR,
- "All tomcat instances failed, no more workers left");
- *is_error = JK_HTTP_SERVER_BUSY;
- rc = JK_FALSE;
+ if (attempt == 0) {
+ int nf;
+ /* Force recovery only on first attempt.
+ * If the second fails, Tomcat is still disconnected.
+ */
+ jk_shm_lock();
+ nf = force_recovery(p->worker, l);
+ jk_shm_unlock();
+
+ if (nf) {
+ jk_log(l, JK_LOG_INFO,
+ "Forcing recovery on first attempt for %d workers",
nf);
+ }
+ else {
+ /* No workers in error state.
+ * Somebody set them all to disabled?
+ */
+ jk_log(l, JK_LOG_ERROR,
+ "All tomcat instances failed, no more workers left
for recovery");
+ *is_error = JK_HTTP_SERVER_BUSY;
+ rc = JK_FALSE;
+ }
+ }
+ else {
+ jk_log(l, JK_LOG_ERROR,
+ "All tomcat instances failed, no more workers left");
+ *is_error = JK_HTTP_SERVER_BUSY;
+ rc = JK_FALSE;
+ }
}
attempt++;
}
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
