core: run many bricks within one glusterfsd process

This patch adds support for multiple brick translator stacks running
in a single brick server process.  This reduces our per-brick memory usage by
approximately 3x, and our appetite for TCP ports even more.  It also creates
potential to avoid process/thread thrashing, and to improve QoS by scheduling
more carefully across the bricks, but realizing that potential will require
further work.

Multiplexing is controlled by the "cluster.brick-multiplex" global option.  By
default it's off, and bricks are started in separate processes as before.  If
multiplexing is enabled, then *compatible* bricks (mostly those with the same
transport options) will be started in the same process.

Change-Id: I45059454e51d6f4cbb29a4953359c09a408695cb
BUG: 1385758
Signed-off-by: Jeff Darcy <jdarcy@redhat.com>
Reviewed-on: https://review.gluster.org/14763
Smoke: Gluster Build System <jenkins@build.gluster.org>
NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
CentOS-regression: Gluster Build System <jenkins@build.gluster.org>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>

1 files changed, 22 insertions, 0 deletions

diff --git a/tests/basic/ec/ec-notify.t b/tests/basic/ec/ec-notify.t
index 586be91bdbe..53290b7c798 100644
--- a/tests/basic/ec/ec-notify.t
+++ b/tests/basic/ec/ec-notify.t
@@ -5,11 +5,26 @@
 
 # This test checks notify part of ec
 
+# We *know* some of these mounts will succeed but not be actually usable
+# (terrible idea IMO), so speed things up and eliminate some noise by
+# overriding this function.
+_GFS () {
+	glusterfs "$@"
+}
+
+ec_up_brick_count () {
+	local bricknum
+	for bricknum in $(seq 0 2); do
+		brick_up_status $V0 $H0 $B0/$V0$bricknum
+	done | grep -E '^1$' | wc -l
+}
+
 cleanup
 TEST glusterd
 TEST pidof glusterd
 TEST $CLI volume create $V0 disperse 3 redundancy 1 $H0:$B0/${V0}{0..2}
 TEST $CLI volume start $V0
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "3" ec_up_brick_count
 
 #First time mount tests.
 # When all the bricks are up, mount should succeed and up-children
@@ -33,6 +48,7 @@ EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0
 
 TEST $CLI volume start $V0
 TEST kill_brick $V0 $H0 $B0/${V0}2
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" ec_up_brick_count
 TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0;
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "2" ec_child_up_count $V0 0
 TEST stat $M0
@@ -40,6 +56,7 @@ EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0
 
 # When only 1 brick is up mount should fail.
 TEST kill_brick $V0 $H0 $B0/${V0}1
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" ec_up_brick_count
 TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0;
 # Wait for 5 seconds even after that up_count should show 1
 sleep 5
@@ -51,28 +68,33 @@ EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0
 # state changes in ec.
 TEST $CLI volume stop $V0
 TEST $CLI volume start $V0 force
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "3" ec_up_brick_count
 TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0;
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "3" ec_child_up_count $V0 0
 TEST touch $M0/a
 
 # kill 1 brick and the up_count should become 2, fops should still succeed
 TEST kill_brick $V0 $H0 $B0/${V0}1
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "2" ec_up_brick_count
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "2" ec_child_up_count $V0 0
 TEST touch $M0/b
 
 # kill one more brick and the up_count should become 1, fops should fail
 TEST kill_brick $V0 $H0 $B0/${V0}2
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" ec_up_brick_count
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" ec_child_up_count $V0 0
 TEST ! touch $M0/c
 
 # kill one more brick and the up_count should become 0, fops should still fail
 TEST kill_brick $V0 $H0 $B0/${V0}0
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "0" ec_up_brick_count
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "0" ec_child_up_count $V0 0
 TEST ! touch $M0/c
 
 # Bring up all the bricks up and see that up_count is 3 and fops are succeeding
 # again.
 TEST $CLI volume start $V0 force
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "3" ec_up_brick_count
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "3" ec_child_up_count $V0 0
 TEST touch $M0/c