tests: move all test-cases into component subdirectories

There are around 300 regression tests, 250 being in tests/bugs. Running
partial set of tests/bugs is not easy because this is a flat directory
with almost all tests inside.

It would be valuable to make partial test/bugs easier, and allow the use
of mulitple build hosts for a single commit, each running a subset of
the tests for a quicker result.

Additional changes made:
- correct the include path for *.rc shell libraries and *.py utils
- make the testcases pass checkpatch
- arequal-checksum in afr/self-heal.t was never executed, now it is
- include.rc now complains loudly if it fails to find env.rc

Change-Id: I26ffd067e9853d3be1fd63b2f37d8aa0fd1b4fea
BUG: 1178685
Reported-by: Emmanuel Dreyfus <manu@netbsd.org>
Reported-by: Atin Mukherjee <amukherj@redhat.com>
URL: http://www.gluster.org/pipermail/gluster-devel/2014-December/043414.html
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Reviewed-on: http://review.gluster.org/9353
Reviewed-by: Kaleb KEITHLEY <kkeithle@redhat.com>
Reviewed-by: Emmanuel Dreyfus <manu@netbsd.org>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>

1 files changed, 91 insertions, 0 deletions

diff --git a/tests/bugs/glusterfs/bug-853690.t b/tests/bugs/glusterfs/bug-853690.t
new file mode 100755
index 00000000000..d81be011438
--- /dev/null
+++ b/tests/bugs/glusterfs/bug-853690.t
@@ -0,0 +1,91 @@
+#!/bin/bash
+#
+# Bug 853690 - Test that short writes do not lead to corruption.
+#
+# Mismanagement of short writes in AFR leads to corruption and immediately
+# detectable split-brain. Write a file to a replica volume using error-gen
+# to cause short writes on one replica.
+#
+# Short writes are also possible during heal. If ignored, the files are marked
+# consistent and silently differ. After reading the file, cause a lookup, wait
+# for self-heal and verify that the afr xattrs do not match.
+#
+########
+
+. $(dirname $0)/../../include.rc
+. $(dirname $0)/../../volume.rc
+
+cleanup;
+
+TEST mkdir -p $B0/test{1,2}
+
+# Our graph is a two brick replica with 100% frequency of short writes on one
+# side of the replica. This guarantees a single write fop leads to an out-of-sync
+# situation.
+cat > $B0/test.vol <<EOF
+volume test-posix-0
+    type storage/posix
+    option directory $B0/test1
+end-volume
+
+volume test-error-0
+    type debug/error-gen
+    option failure 100
+    option enable writev
+    option error-no GF_ERROR_SHORT_WRITE
+    subvolumes test-posix-0
+end-volume
+
+volume test-locks-0
+    type features/locks
+    subvolumes test-error-0
+end-volume
+
+volume test-posix-1
+    type storage/posix
+    option directory $B0/test2
+end-volume
+
+volume test-locks-1
+    type features/locks
+    subvolumes test-posix-1
+end-volume
+
+volume test-replicate-0
+    type cluster/replicate
+    option background-self-heal-count 0
+    subvolumes test-locks-0 test-locks-1
+end-volume
+EOF
+
+TEST glusterd
+
+TEST glusterfs --volfile=$B0/test.vol --attribute-timeout=0 --entry-timeout=0 $M0
+
+# Send a single write, guaranteed to be short on one replica, and attempt to
+# read the data back. Failure to detect the short write results in different
+# file sizes and immediate split-brain (EIO).
+TEST dd if=/dev/zero of=$M0/file bs=128k count=1
+TEST dd if=$M0/file of=/dev/null bs=128k count=1
+########
+#
+# Test self-heal with short writes...
+#
+########
+
+# Cause a lookup and wait a few seconds for posterity. This self-heal also fails
+# due to a short write.
+TEST ls $M0/file
+# Verify the attributes on the healthy replica do not reflect consistency with
+# the other replica.
+xa=`getfattr -n trusted.afr.test-locks-0 -e hex $B0/test2/file 2>&1 | grep = | cut -f2 -d=`
+EXPECT_NOT 0x000000000000000000000000 echo $xa
+
+TEST rm -f $M0/file
+EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0
+
+rm -f $B0/test.vol
+rm -rf $B0/test1 $B0/test2
+
+cleanup;
+