ctdb/web/testing.html

   1 <!--#set var="TITLE" value="CTDB Testing" -->
   2 <!--#include virtual="header.html" -->
   3
   4 <H2 align="center">Starting and testing CTDB</h2>
   5
   6 The CTDB log is in /var/log/log.ctdb so look in this file if something
   7 did not start correctly.<p>
   8
   9 You can ensure that ctdb is running on all nodes using
  10 <pre>
  11   onnode all service ctdb start
  12 </pre>
  13 Verify that the CTDB daemon started properly. There should normally be at least 2 processes started for CTDB, one for the main daemon and one for the recovery daemon.
  14 <pre>
  15   onnode all pidof ctdbd
  16 </pre>
  17
  18 Once all CTDB nodes have started, verify that they are correctly
  19 talking to each other.<p>
  20
  21 There should be one TCP connection from the private ip address on each
  22 node to TCP port 4379 on each of the other nodes in the cluster.
  23 <pre>
  24   onnode all netstat -tn | grep 4379
  25 </pre>
  26
  27
  28 <h2>Automatically restarting CTDB</h2>
  29
  30 If you wish to cope with software faults in ctdb, or want ctdb to
  31 automatically restart when an administration kills it, then you may
  32 wish to add a cron entry for root like this:
  33
  34 <pre>
  35  * * * * * /etc/init.d/ctdb cron > /dev/null 2>&1
  36 </pre>
  37
  38
  39 <h2>Testing CTDB</h2>
  40
  41 Once your cluster is up and running, you may wish to know how to test that it is functioning correctly. The following tests may help with that
  42
  43 <h3>The ctdb tool</h3>
  44
  45 The ctdb package comes with a utility called ctdb that can be used to
  46 view the behaviour of the ctdb cluster.<p>
  47
  48 If you run it with no options it will provide some terse usage information. The most commonly used commands are:
  49 <pre>
  50  ctdb status
  51  ctdb ip
  52  ctdb ping
  53 </pre>
  54
  55 <h3>ctdb status</h3>
  56
  57 The status command provides basic information about the cluster and the status of the nodes. when you run it you will get some output like:
  58
  59 <pre>
  60 <strong>Number of nodes:4
  61 vnn:0 10.1.1.1       OK (THIS NODE)
  62 vnn:1 10.1.1.2       OK
  63 vnn:2 10.1.1.3       OK
  64 vnn:3 10.1.1.4       OK</strong>
  65 Generation:1362079228
  66 Size:4
  67 hash:0 lmaster:0
  68 hash:1 lmaster:1
  69 hash:2 lmaster:2
  70 hash:3 lmaster:3
  71 <strong>Recovery mode:NORMAL (0)</strong>
  72 Recovery master:0
  73 </pre>
  74
  75 The important parts are in bold. This tells us that all 4 nodes are in
  76 a healthy state.<p>
  77
  78 It also tells us that recovery mode is normal, which means that the
  79 cluster has finished a recovery and is running in a normal fully
  80 operational state.<p>
  81
  82 Recovery state will briefly change to "RECOVERY" when there ahs been a
  83 node failure or something is wrong with the cluster.<p>
  84
  85 If the cluster remains in RECOVERY state for very long (many seconds)
  86 there might be something wrong with the configuration. See
  87 /var/log/log.ctdb.
  88
  89 <h3>ctdb ip</h3>
  90
  91 This command prints the current status of the public ip addresses and which physical node is currently serving that ip.
  92
  93 <pre>
  94 Number of nodes:4
  95 192.168.1.1         0
  96 192.168.1.2         1
  97 192.168.2.1         2
  98 192.168.2.1         3
  99 </pre>
 100
 101 <h3>ctdb ping</h3>
 102 this command tries to "ping" each of the CTDB daemons in the cluster.
 103 <pre>
 104   ctdb ping -n all
 105
 106   response from 0 time=0.000050 sec  (13 clients)
 107   response from 1 time=0.000154 sec  (27 clients)
 108   response from 2 time=0.000114 sec  (17 clients)
 109   response from 3 time=0.000115 sec  (59 clients)
 110 </pre>
 111
 112 <!--#include virtual="footer.html" -->