[Checkins] SVN: gocept.zeoraid/trunk/ Improve controller documentation. Especially document real parameters of the

[Christian Theune]

Log message for revision 105641:
  Improve controller documentation. Especially document real parameters of the
  reload command.
  

Changed:
  U   gocept.zeoraid/trunk/doc/CMDLINE.txt
  U   gocept.zeoraid/trunk/src/gocept/zeoraid/scripts/controller.py

-=-
Modified: gocept.zeoraid/trunk/doc/CMDLINE.txt
===================================================================
--- gocept.zeoraid/trunk/doc/CMDLINE.txt	2009-11-14 09:00:33 UTC (rev 105640)
+++ gocept.zeoraid/trunk/doc/CMDLINE.txt	2009-11-14 09:14:46 UTC (rev 105641)
@@ -22,14 +22,49 @@
 
     disable <storage> -- Disable a storage
 
-    reload </path/to/zeo.conf> -- Reload a specified zeo.conf file
+    reload -- Reload this RAID storage's configuration from the server's
+              zeo.conf.
 
+Defaults
+--------
+
+The defaults can be provided in a modified version by using the
+gocept.zeoraid:server recipe for generating a ZEO server with RAID
+configuration and installing a console script with matching defaults.
+
+The defaults listed above are used when the console script is installed
+without the buildout recipe.
+
+
 recover
 -------
-Multiple recoveries can't happen in parallel (see KNOWNBUGS.txt for more info).
 
+Multiple recoveries can't happen in parallel (see KNOWNBUGS.txt for more
+information).
+
+
 reload
 ------
-The reload command takes a zeo.conf file and compares the names of the storages described there and the names of the storages currently loaded to see if any storage needs to be added and/or removed. A new storage is added in the degraded state, so you need to manually 'recover' it, while a removed storage is 'disabled', not really removed. 
 
-Example: Let suppose we have a ZEORaid server connected with three storages called '1', '2' and '3' in the optimal state. Now we call 'reload' on a modified zeo.conf file that have the storages '1', '3', '4' and '5' - the storage '2' is disabled and the storages '4' and '5' are added as degraded storages. We now must 'recover' them, one at a time, before using them.
\ No newline at end of file
+The reload command re-reads the ZEO server's configuration file and for the
+RAID storage given checks the list of back-end storages for additions and
+removals.
+
+New back-ends are added to the RAID in `degraded` state and have to be
+manually recovered.
+
+Back-ends that are missing from the new configuration are `disabled` but
+remain in the RAID.
+
+Example:
+
+Suppose we have a ZEORaid server connected to three back-ends named '1', '2'
+and '3'. All of them are currently in the `optimal` state.
+
+Now we modify the ZEO server's configuration file so that it lists back-ends
+'1', '3', '4', and '5': storage '2' is removed and storages '4' and '5' are
+added.
+
+When calling `reload` storage '2' will become disabled and storages '4' and
+'5' are shown as degraded. We have to run 'reload' for each of them to get
+back to `optimal`.

Modified: gocept.zeoraid/trunk/src/gocept/zeoraid/scripts/controller.py
===================================================================
--- gocept.zeoraid/trunk/src/gocept/zeoraid/scripts/controller.py	2009-11-14 09:00:33 UTC (rev 105640)
+++ gocept.zeoraid/trunk/src/gocept/zeoraid/scripts/controller.py	2009-11-14 09:14:46 UTC (rev 105641)
@@ -132,5 +132,6 @@
         result = NAGIOS_CRITICAL
     sys.exit(result)
 
+
 if __name__ == '__main__':
     main()