Signed-off-by: Valerio Pachera <[email protected]> --- doc/stop_and_restart_cluster.rst | 39 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) create mode 100644 doc/stop_and_restart_cluster.rst
diff --git a/doc/stop_and_restart_cluster.rst b/doc/stop_and_restart_cluster.rst new file mode 100644 index 0000000..07571e7 --- /dev/null +++ b/doc/stop_and_restart_cluster.rst @@ -0,0 +1,39 @@ +Stop And Restart +================ + +Stop The Cluster +**************** + +It's very important to stop the cluster in the right way to avoid corruption +of vdi. + +First of all, shut down all virtual machines on every node. +Once done, run + +:: + + # dog cluster shutdown + +This will kill all sheep daemon on all nodes, without start/interrupt any +rebuilding. +Now it's safe to shut down the hosts. + +**Remember: +DO NOT shut down any of node before deactivating the cluster! +DO NOT shut down / kill a node after the other.** + +Restart The Cluster +******************* + +Run sheep daemon on every node, on the same directory as before. +Don't wary, no rebuild will start after the second or third node goes on. +If sheepdog has been stopped with 4 nodes alive, the cluster will not start +till all 4 nodes will be on line. + +E.g. start sheep on the first node and run + +:: + + dog cluster info + + Cluster status: Waiting for other nodes to join cluster -- 1.7.10.4 -- sheepdog mailing list [email protected] http://lists.wpkg.org/mailman/listinfo/sheepdog
