<div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">What makes you think so?<br>
Anything about "full sync" in the logs?<br>
<div class="im"><br></div></blockquote><div><br></div><div><br></div><div><div>Nov 18 09:06:08 seaspray kernel: [159990.699555] block drbd0: Writing the whole </div><div>bitmap, full sync required after drbd_sync_handshake.</div>
<div>Nov 18 09:06:08 seaspray kernel: [159990.699632] block drbd1: drbd_sync_handshak</div><div>e:</div></div><div><br></div><div><br></div><div>and then cat /proc/drbd show the whole full resync process.</div><div><br></div>
<div>I just test again, stop the secondary, wait for a while, this time mysql is still writing new data to the primary node- restart the secondary. It is not a full resync now. Weird.</div><div><br></div><div>Repeat the test several times with consistent manner - not really quite sure why it requiers full resync in the morning.</div>
<div><br></div><div>Thanks,</div><div><br></div><div><br></div></div><br clear="all"><div><br></div>-- <br>Steve Kieu<br>