summaryrefslogtreecommitdiffstats
path: root/docs/testing/user/testspecification/highavailability/index.rst
diff options
context:
space:
mode:
Diffstat (limited to 'docs/testing/user/testspecification/highavailability/index.rst')
-rw-r--r--docs/testing/user/testspecification/highavailability/index.rst7
1 files changed, 4 insertions, 3 deletions
diff --git a/docs/testing/user/testspecification/highavailability/index.rst b/docs/testing/user/testspecification/highavailability/index.rst
index e024b1fc..dd98ba94 100644
--- a/docs/testing/user/testspecification/highavailability/index.rst
+++ b/docs/testing/user/testspecification/highavailability/index.rst
@@ -810,9 +810,10 @@ For second monitor is used a process monitor and the main purpose is to watch
whether the database processes on the host node are killed properly.
Therefore, in this test case, there are two metrics:
-- service_outage_time, which indicates the maximum outage time (seconds)
+
+* service_outage_time, which indicates the maximum outage time (seconds)
of the specified OpenStack command request
-- process_recover_time, which indicates the maximum time (seconds) from the
+* process_recover_time, which indicates the maximum time (seconds) from the
process being killed to recovered
Test execution
@@ -1056,7 +1057,7 @@ Process recovery is verified by checking the existence of processes of
Test execution
''''''''''''''
* Test action 1: Two host VMs are booted, these two hosts are in two different
-networks, the networks are connected by a virtual router.
+ networks, the networks are connected by a virtual router.
* Test action 2: Start monitors: each monitor will run with independently process.
The monitor info will be collected.
* Test action 3: Do attacker: Connect the host through SSH, and then execute the kill