Add Ceph report from Willem Jan Withagen <wjw@digiware.nl>.

This commit is contained in:
Warren Block 2016-04-13 16:51:56 +00:00
parent d5208c2ded
commit bafbad80d2
Notes: svn2git 2020-12-08 03:00:23 +00:00
svn path=/head/; revision=48612

View file

@ -2013,4 +2013,202 @@
</task>
</help>
</project>
<project cat='proj'>
<title>Ceph on FreeBSD</title>
<contact>
<person>
<name>
<given>Willem Jan</given>
<common>Withagen</common>
</name>
<email>wjw@digiware.nl</email>
</person>
</contact>
<links>
<url href="http://ceph.com">Ceph main site</url>
<url href="https://github.com/ceph/ceph">Main repository</url>
<url href="https://github.com/wjwithagen/ceph">My Fork</url>
<url href="https://github.com/ceph/ceph/pull/7573">The git PULL with all changes</url>
</links>
<body>
<p>Ceph is a distributed object store and file system designed
to provide excellent performance, reliability and
scalability.</p>
<ul>
<li>
<p>Object Storage</p>
<p>Ceph provides seamless access to objects using native
language bindings or radosgw, a REST interface that is
compatible with applications written for S3 and Swift.</p>
</li>
<li>
<p>Block Storage</p>
<p>Ceph's RADOS Block Device (RBD) provides access to block
device images that are striped and replicated across the
entire storage cluster.</p>
</li>
<li>
<p>File System</p>
<p>Ceph provides a POSIX-compliant network file system that
aims for high performance, large data storage, and maximum
compatibility with legacy applications.</p>
</li>
</ul>
<p>I started looking into Ceph, because the HAST solution with
CARP and <tt>ggate</tt> did not really do what I wanted. But
I am aiming for running a Ceph storage cluster of storage
nodes that are running ZFS. The end station would be running
<tt>bhyve</tt> on RBD disk that are stored in Ceph.</p>
<p>The FreeBSD build will build most of the tools in Ceph. Note
that the RBD-dependent items will not work since FreeBSD does
not have RBD yet.</p>
<p>Build Prerequisites</p>
<p>Compiling and building Ceph is tested on 11-CURRENT. It uses
the CLANG toolset taht is available, which needs to be at
least 3.7. Clang 3.4 (on 10.2-STABLE) does not have all the
required capabilites to compile everything.</p>
<p>This setup will get things running for FreeBSD:</p>
<ul>
<li>
<p>Install bash and link it in <tt>/bin</tt> (requires root
privileges):</p>
<p><tt>sudo pkg install bash</tt></p>
<p><tt>sudo ln -s /usr/local/bin/bash /bin/bash</tt></p>
</li>
<li>
<p>Building Ceph</p>
<p><tt>./do_freebsd.sh</tt></p>
</li>
</ul>
<p>Parts Not Yet Included</p>
<ul>
<li>
<p>RBD</p>
<p>Rados Block Devices is implemented in the Linux kernel.
It seems that there used to be a userspace implementation
first. And perhaps <tt>ggated</tt> could be used as a
template since it does some of the same functions, other
than just between two disks. And it has a userspace
counterpart.</p>
</li>
<li>
<p>BlueStore</p>
<p>FreeBSD and Linux have a different AIO API, and that
needs to be made compatible. Next to that is the
discussion in FreeBSD about <tt>aio_cancel</tt> not
working for all device types.</p>
</li>
<li>
<p>CephFS</p>
<p>Cython tries to access an internal field in dirent which
does not compile.</p>
</li>
</ul>
<p>Tests that verify the correct working of the above are also
excluded from the test set.</p>
<p>Tests Not Yet Included</p>
<ul>
<li>
<p><tt>ceph-detect-init/run-tox.sh</tt></p>
<p>Because the current implementation does not know anything
about FreeBSD rc-init.</p>
</li>
<li>
<p>Tests that make use of nosestests</p>
<p>Calling these does not really work since
<tt>nosetests</tt> is not in <tt>/usr/bin</tt>, and
calling through <tt>/usr/bin/env/nosetests</tt> does not
work on FreeBSD.</p>
</li>
<li>
<p><tt>test/pybind/test_ceph_argparse.py</tt></p>
</li>
<li>
<p><tt>test/pybind/test_ceph_daemon.py</tt></p>
</li>
</ul>
<p>Things To Investigate</p>
<ul>
<li>
<p><tt>ceph-{osd,mon}</tt> need two signals before they
actually terminate.</p>
</li>
<li>
<p><tt>ceph_erasure_code --debug-osd 20 --plugin_exists jerasure</tt>
crashes due to SIGSEGV. This is a pointer reference that
gets modified outside the regular programflow. Probably
due to a programming error but perhaps wrong mixing and
matching of many libraries.</p>
</li>
</ul>
</body>
<help>
<task>
<p>Current and foremost task is to get the test set to
complete without errors. This includes fixing several
coredumps.</p>
<p>Run integration tests to see if the FreeBSD daemons will
work with a Linux Ceph platform.</p>
</task>
<task>
<p>Get the Python tests that are currently excluded to work,
and test OKE.</p>
</task>
<task>
<p>Compile and test the user space RBD (Rados Block
Device).</p>
<p>Investigate and see if an in-kernel RBD device could be
developed a la <tt>ggate</tt>.</p>
</task>
<task>
<p>Integrate the FreeBSD <tt>/etc/rc.d</tt> init scripts in
the Ceph stack for testing and running Ceph on production
machines.</p>
</task>
</help>
</project>
</report>