<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">some results of online discussion.<div><br></div><div>Johann worry about journal checksumming (to avoid overhead in that area) or other metadata checks on </div><div>client store uncommitted data in own queue for a replay (if need), but in case journal checksum errors on write client will able to resend failed request via special protocol extension (don't exist now).</div><div>same situation if we need a T10 support for these data we will be need trigger a failover or change a wire protocol to resend a failed request.</div><div>failover is easily.</div><div><br></div><div><br><div><div>On Apr 24, 2012, at 14:53, Johann Lombardi wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div class="gmail_extra">On Tue, Apr 24, 2012 at 9:44 PM, Alexey Lyashkov <span dir="ltr"><<a href="mailto:alexey_lyashkov@xyratex.com" target="_blank">alexey_lyashkov@xyratex.com</a>></span> wrote:<br><div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word">what you mean as metadata blocks? extent's on OST or MDT disk io traffic ?</div></blockquote>
<div><br>Yes, anything that goes to the journal (inode, block bitmap, ...) which isn't user data.<br><br>Johann<br></div></div></div>
</blockquote></div><br></div></body></html>