From d7901746f498f82fec8f972f9ce62a2003ad0936 Mon Sep 17 00:00:00 2001
From: sageweil <sageweil@29311d96-e01e-0410-9327-a35deaab8ce9>
Date: Sat, 2 Dec 2006 02:28:49 +0000
Subject: [PATCH] more on mpi

git-svn-id: https://ceph.svn.sf.net/svnroot/ceph@975 29311d96-e01e-0410-9327-a35deaab8ce9
---
 web/source.body | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/web/source.body b/web/source.body
index 9e059b3729849..983b001c70742 100644
--- a/web/source.body
+++ b/web/source.body
@@ -103,9 +103,11 @@ mpd &           # for a single host
 mpiboot -n 10   # for multiple hosts (see MPICH docs)
 make newsyn && ./newsyn --mkfs --nummds 2 --numosd 6 --numclient 20 --syn writefile 100 16384
 </pre>
+			You will probably want to make <tt>dev/osdall</tt> a symlink to some block device that exists on every node you're starting an OSD on.  Otherwise, you'll need a symlink (for "block device" file) for each osd.
+
 			If you want to mount a distributed FS (instead of generating a synthetic workload), try
 <pre>
-make newsyn && ./newsyn --mkfs --nummds 2 --numosd 6 --numclient 0
+make newsyn && ./newsyn --mkfs --nummds 2 --numosd 6 --numclient 0     # 0 clients, just mds and osds
 # in another terminal,
 mkdir mnt
 make cfuse && ./cfuse mnt
@@ -113,7 +115,7 @@ make cfuse && ./cfuse mnt
 ls mnt 
 touch mnt/asdf   # etc
 </pre>
-			Currently, when the last client (<tt>cfuse</tt> instance, in this case) shuts down, the whole thing will shut down.  Assuming things shut down cleanly, you should be able to start things up again without the <tt>--mkfs--</tt> flag and recovery the prior file system state.
+			Currently, when the last client (<tt>cfuse</tt> instance, in this case) shuts down, the whole thing will shut down.  Assuming things shut down cleanly, you should be able to start things up again without the <tt>--mkfs</tt> flag and recover the prior file system state.
 		</div>
 
 		<h4>Structure</h4>
-- 
2.39.5