<html><head></head><body><div class="ydpf292fe38yahoo-style-wrap" style="font-family:courier new, courier, monaco, monospace, sans-serif;font-size:16px;"><div></div>
<div><span style="color: rgb(38, 40, 42); font-family: Helvetica Neue, Helvetica, Arial, sans-serif; font-size: 13px;">>What I do have is an extra SSD disk per server which I plan to
use as LVM cache for the bricks (Maybe just 1 disk, maybe 2 >with
SW RAID 1). I still need to test how LVM / gluster are going to
handle the failure of the cache disk.</span><br></div></div><div id="ydp6f0a849cyahoo_quoted_0676875783" class="ydp6f0a849cyahoo_quoted"><div style="font-family:'Helvetica Neue', Helvetica, Arial, sans-serif;font-size:13px;color:#26282a;"><div><div id="ydp6f0a849cyiv7302767365">
<div><br></div><div>Are you planing to use LVM cache for reads (writethrough) only or both read + writes (writeback) ? </div><div><br></div><div>If you pick the writeback - which means that you first write to the SSDs and only then push the data to the slow HDDs , then you need a raid1 for the SSD-based LVM cache , or a pure replica3 cause if you loose 1 SSD - the whole brick is down and you need to sync from scratch - and with the size of data you have - this will not be so nice.</div><div><br></div><div>Note: Don't forget that thin LVM is the only way that gluster can use snapshots.</div><div><br></div><div>Best Regards,</div><div>Strahil Nikolov</div></div></div>
</div>
</div></body></html>