From: Sage Weil <sage@redhat.com>
Date: Tue, 31 Jul 2018 14:38:39 +0000 (-0500)
Subject: doc/mgr/devicehealth: document devicehealth module
X-Git-Tag: v14.0.1~723^2
X-Git-Url: http://git-server-git.apps.pok.os.sepia.ceph.com/?a=commitdiff_plain;h=f09a87f9025fca332697f66be4a3c889fe80ccc9;p=ceph.git

doc/mgr/devicehealth: document devicehealth module

Signed-off-by: Sage Weil <sage@redhat.com>
---

diff --git a/doc/mgr/devicehealth.rst b/doc/mgr/devicehealth.rst
new file mode 100644
index 000000000000..5e0d0012192b
--- /dev/null
+++ b/doc/mgr/devicehealth.rst
@@ -0,0 +1,52 @@
+Devicehealth plugin
+===================
+
+The *devicehealth* plugin includes code to manage physical devices
+that back Ceph daemons (e.g., OSDs).  This includes scraping health
+metrics (e.g., SMART) and responding to health metrics by migrating
+data away from failing devices.
+
+Enabling
+--------
+
+The *devicehealth* module is enabled with::
+
+  ceph mgr module enable devicehealth
+
+(It is enabled by default.)
+
+Scraping
+--------
+
+Health metrics can be scraped from all devices with::
+
+  ceph device scrape-health-metrics
+
+A single device can be scraped with::
+
+  ceph device scrape-health-metrics <device-id>
+
+Or a single daemon's devices can be scraped with::
+
+  ceph device scrape-daemon-health-metrics <who>
+
+
+Health monitoring
+-----------------
+
+By default, the devicehealth module wakes up periodically and checks
+the health of all devices in the system.  This will raise health
+alerts if devices are expected to fail soon.  This can be disabled by
+turning off the ``mgr/devicehealth/enable_monitoring`` option.
+
+The ``mgr/devicehealth/warn_threshold`` controls how soon an expected
+device failure must be before we generate a health warning.
+
+If the ``mgr/devicehealth/self_heal`` option is enabled (it is by
+default), then for devices that are expected to fail soon the module
+will automatically migrate data away from them by marking the devices
+"out".
+
+The ``mgr/devicehealth/mark_out_threshold`` controls how soon an
+expected device failure must be before we automatically mark an osd
+"out".
diff --git a/doc/mgr/index.rst b/doc/mgr/index.rst
index ea8c9d48ca1b..e640292f40f4 100644
--- a/doc/mgr/index.rst
+++ b/doc/mgr/index.rst
@@ -39,3 +39,4 @@ sensible.
     Telemetry plugin <telemetry>
     Iostat plugin <iostat>
     Crash plugin <crash>
+    Devicehealth plugin <devicehealth>