ceph/doc/mgr/administrator.rst

   1
   2 ceph-mgr administrator's guide
   3 ==============================
   4
   5 Manual setup
   6 ------------
   7
   8 Usually, you would set up a ceph-mgr daemon using a tool such
   9 as ceph-ansible.  These instructions describe how to set up
  10 a ceph-mgr daemon manually.
  11
  12 First, create an authentication key for your daemon::
  13
  14     ceph auth get-or-create mgr.$name mon 'allow profile mgr' osd 'allow *' mds 'allow *'
  15
  16 Place that key into ``mgr data`` path, which for a cluster "ceph"
  17 and mgr $name "foo" would be ``/var/lib/ceph/mgr/ceph-foo``.
  18
  19 Start the ceph-mgr daemon::
  20
  21     ceph-mgr -i $name
  22
  23 Check that the mgr has come up by looking at the output
  24 of ``ceph status``, which should now include a mgr status line::
  25
  26     mgr active: $name
  27
  28 Client authentication
  29 ---------------------
  30 The manager is a new daemon which requires new CephX capabilities. If you upgrade
  31 a cluster from an old version of Ceph, or use the default install/deploy tools,
  32 your admin client should get this capability automatically. If you use tooling from
  33 elsewhere, you may get EACCES errors when invoking certain ceph cluster commands.
  34 To fix that, add a "mgr allow *" stanza to your client's cephx capabilities by
  35 `Modifying User Capabilities`_.
  36
  37 High availability
  38 -----------------
  39
  40 In general, you should set up a ceph-mgr on each of the hosts
  41 running a ceph-mon daemon to achieve the same level of availability.
  42
  43 By default, whichever ceph-mgr instance comes up first will be made
  44 active by the monitors, and the others will be standbys.  There is
  45 no requirement for quorum among the ceph-mgr daemons.
  46
  47 If the active daemon fails to send a beacon to the monitors for
  48 more than ``mon mgr beacon grace`` (default 30s), then it will be replaced
  49 by a standby.
  50
  51 If you want to pre-empt failover, you can explicitly mark a ceph-mgr
  52 daemon as failed using ``ceph mgr fail <mgr name>``.
  53
  54 Calling module commands
  55 -----------------------
  56
  57 Where a module implements command line hooks, using the Ceph CLI's
  58 ``tell`` command to call them like this::
  59
  60     ceph tell mgr <command | help>
  61
  62 Note that it is not necessary to address a particular mgr instance,
  63 simply ``mgr`` will pick the current active daemon.
  64
  65 Use the ``help`` command to get a list of available commands from all
  66 modules.
  67
  68 Configuration
  69 -------------
  70
  71 OPTION(mgr_module_path, OPT_STR, CEPH_PKGLIBDIR "/mgr") // where to load python modules from
  72
  73 ``mgr module path``
  74
  75 :Description: Path to load modules from
  76 :Type: String
  77 :Default: ``"<library dir>/mgr"``
  78
  79 ``mgr modules``
  80
  81 :Description: List of python modules to load
  82 :Type: String
  83 :Default: ``"rest"`` (Load the REST API module only)
  84
  85 ``mgr data``
  86
  87 :Description: Path to load daemon data (such as keyring)
  88 :Type: String
  89 :Default: ``"/var/lib/ceph/mgr/$cluster-$id"``
  90
  91 ``mgr tick period``
  92
  93 :Description: How many seconds between mgr beacons to monitors, and other
  94               periodic checks.
  95 :Type: Integer
  96 :Default: ``5``
  97
  98 ``mon mgr beacon grace``
  99
 100 :Description: How long after last beacon should a mgr be considered failed
 101 :Type: Integer
 102 :Default: ``30``
 103
 104 .. _Modifying User Capabilities: ../rados/operations/user-management#modify-user-capabilities