ceph/doc/install/ceph-deploy/upgrading-ceph.rst

   1 ================
   2  Upgrading Ceph
   3 ================
   4
   5 Each release of Ceph may have additional steps. Refer to the `release notes
   6 document of your release`_ to identify release-specific procedures for your
   7 cluster before using the upgrade procedures.
   8
   9
  10 Summary
  11 =======
  12
  13 You can upgrade daemons in your Ceph cluster while the cluster is online and in
  14 service! Certain types of daemons depend upon others. For example, Ceph Metadata
  15 Servers and Ceph Object Gateways depend upon Ceph Monitors and Ceph OSD Daemons.
  16 We recommend upgrading in this order:
  17
  18 #. `Ceph Deploy`_
  19 #. Ceph Monitors
  20 #. Ceph OSD Daemons
  21 #. Ceph Metadata Servers
  22 #. Ceph Object Gateways
  23
  24 As a general rule, we recommend upgrading all the daemons of a specific type
  25 (e.g., all ``ceph-mon`` daemons, all ``ceph-osd`` daemons, etc.) to ensure that
  26 they are all on the same release. We also recommend that you upgrade all the
  27 daemons in your cluster before you try to exercise new functionality in a
  28 release.
  29
  30 The `Upgrade Procedures`_ are relatively simple, but do look at the `release
  31 notes document of your release`_ before upgrading. The basic process involves
  32 three steps:
  33
  34 #. Use ``ceph-deploy`` on your admin node to upgrade the packages for
  35    multiple hosts (using the ``ceph-deploy install`` command), or login to each
  36    host and upgrade the Ceph package `using your distro's package manager`_.
  37    For example, when `Upgrading Monitors`_, the ``ceph-deploy`` syntax might
  38    look like this::
  39
  40         ceph-deploy install --release {release-name} ceph-node1[ ceph-node2]
  41         ceph-deploy install --release firefly mon1 mon2 mon3
  42
  43    **Note:** The ``ceph-deploy install`` command will upgrade the packages
  44    in the specified node(s) from the old release to the release you specify.
  45    There is no ``ceph-deploy upgrade`` command.
  46
  47 #. Login in to each Ceph node and restart each Ceph daemon.
  48    See `Operating a Cluster`_ for details.
  49
  50 #. Ensure your cluster is healthy. See `Monitoring a Cluster`_ for details.
  51
  52 .. important:: Once you upgrade a daemon, you cannot downgrade it.
  53
  54
  55 Ceph Deploy
  56 ===========
  57
  58 Before upgrading Ceph daemons, upgrade the ``ceph-deploy`` tool. ::
  59
  60         sudo pip install -U ceph-deploy
  61
  62 Or::
  63
  64         sudo apt-get install ceph-deploy
  65
  66 Or::
  67
  68         sudo yum install ceph-deploy python-pushy
  69
  70
  71 Upgrade Procedures
  72 ==================
  73
  74 The following sections describe the upgrade process.
  75
  76 .. important:: Each release of Ceph may have some additional steps. Refer to
  77    the `release notes document of your release`_ for details **BEFORE** you
  78    begin upgrading daemons.
  79
  80
  81 Upgrading Monitors
  82 ------------------
  83
  84 To upgrade monitors, perform the following steps:
  85
  86 #. Upgrade the Ceph package for each daemon instance.
  87
  88    You may use ``ceph-deploy`` to address all monitor nodes at once.
  89    For example::
  90
  91         ceph-deploy install --release {release-name} ceph-node1[ ceph-node2]
  92         ceph-deploy install --release hammer mon1 mon2 mon3
  93
  94    You may also use the package manager for your Linux distribution on
  95    each individual node. To upgrade packages manually on each Debian/Ubuntu
  96    host, perform the following steps::
  97
  98         ssh {mon-host}
  99         sudo apt-get update && sudo apt-get install ceph
 100
 101    On CentOS/Red Hat hosts, perform the following steps::
 102
 103         ssh {mon-host}
 104         sudo yum update && sudo yum install ceph
 105
 106
 107 #. Restart each monitor. For Ubuntu distributions, use::
 108
 109         sudo systemctl restart ceph-mon@{hostname}.service
 110
 111    For CentOS/Red Hat/Debian distributions, use::
 112
 113         sudo /etc/init.d/ceph restart {mon-id}
 114
 115    For CentOS/Red Hat distributions deployed with ``ceph-deploy``,
 116    the monitor ID is usually ``mon.{hostname}``.
 117
 118 #. Ensure each monitor has rejoined the quorum::
 119
 120         ceph mon stat
 121
 122 Ensure that you have completed the upgrade cycle for all of your Ceph Monitors.
 123
 124
 125 Upgrading an OSD
 126 ----------------
 127
 128 To upgrade a Ceph OSD Daemon, perform the following steps:
 129
 130 #. Upgrade the Ceph OSD Daemon package.
 131
 132    You may use ``ceph-deploy`` to address all Ceph OSD Daemon nodes at
 133    once. For example::
 134
 135         ceph-deploy install --release {release-name} ceph-node1[ ceph-node2]
 136         ceph-deploy install --release hammer osd1 osd2 osd3
 137
 138    You may also use the package manager on each node to upgrade packages
 139    `using your distro's package manager`_. For Debian/Ubuntu hosts, perform the
 140    following steps on each host::
 141
 142         ssh {osd-host}
 143         sudo apt-get update && sudo apt-get install ceph
 144
 145    For CentOS/Red Hat hosts, perform the following steps::
 146
 147         ssh {osd-host}
 148         sudo yum update && sudo yum install ceph
 149
 150
 151 #. Restart the OSD, where ``N`` is the OSD number. For Ubuntu, use::
 152
 153         sudo systemctl restart ceph-osd@{N}.service
 154
 155    For multiple OSDs on a host, you may restart all of them with systemd. ::
 156
 157         sudo systemctl restart ceph-osd
 158
 159    For CentOS/Red Hat/Debian distributions, use::
 160
 161         sudo /etc/init.d/ceph restart N
 162
 163
 164 #. Ensure each upgraded Ceph OSD Daemon has rejoined the cluster::
 165
 166         ceph osd stat
 167
 168 Ensure that you have completed the upgrade cycle for all of your
 169 Ceph OSD Daemons.
 170
 171
 172 Upgrading a Metadata Server
 173 ---------------------------
 174
 175 To upgrade a Ceph Metadata Server, perform the following steps:
 176
 177 #. Upgrade the Ceph Metadata Server package. You may use ``ceph-deploy`` to
 178    address all Ceph Metadata Server nodes at once, or use the package manager
 179    on each node. For example::
 180
 181         ceph-deploy install --release {release-name} ceph-node1
 182         ceph-deploy install --release hammer mds1
 183
 184    To upgrade packages manually, perform the following steps on each
 185    Debian/Ubuntu host::
 186
 187         ssh {mon-host}
 188         sudo apt-get update && sudo apt-get install ceph-mds
 189
 190    Or the following steps on CentOS/Red Hat hosts::
 191
 192         ssh {mon-host}
 193         sudo yum update && sudo yum install ceph-mds
 194
 195
 196 #. Restart the metadata server. For Ubuntu, use::
 197
 198         sudo systemctl restart ceph-mds@{hostname}.service
 199
 200    For CentOS/Red Hat/Debian distributions, use::
 201
 202         sudo /etc/init.d/ceph restart mds.{hostname}
 203
 204    For clusters deployed with ``ceph-deploy``, the name is usually either
 205    the name you specified on creation or the hostname.
 206
 207 #. Ensure the metadata server is up and running::
 208
 209         ceph mds stat
 210
 211
 212 Upgrading a Client
 213 ------------------
 214
 215 Once you have upgraded the packages and restarted daemons on your Ceph
 216 cluster, we recommend upgrading ``ceph-common`` and client libraries
 217 (``librbd1`` and ``librados2``) on your client nodes too.
 218
 219 #. Upgrade the package::
 220
 221         ssh {client-host}
 222         apt-get update && sudo apt-get install ceph-common librados2 librbd1 python-rados python-rbd
 223
 224 #. Ensure that you have the latest version::
 225
 226         ceph --version
 227
 228 If you do not have the latest version, you may need to uninstall, auto remove
 229 dependencies and reinstall.
 230
 231
 232 .. _using your distro's package manager: ../install-storage-cluster/
 233 .. _Operating a Cluster: ../../rados/operations/operating
 234 .. _Monitoring a Cluster: ../../rados/operations/monitoring
 235 .. _release notes document of your release: ../../releases