Hey, all.<br><br>I'm currently trying out GlusterFS 3.3.<br>I've got two servers and four clients, all on separate boxes.<br>I've got a Distributed-Replicated volume with 4 bricks, two from each server,<br>and I'm using the FUSE client.<br>
<br>I was trying out failover, currently testing for reads.<br>I was reading a big file, using iftop to see which server was actually being read from.<br>I put up an iptables firewall on the active server, and the client hung, as expected.<br>
After some time, I guess about 30 seconds, the failover kicked in and the operations<br>went to the other server, resuming the read.<br><br>However, after going to the other server, throughput went from ~100MB/s to ~25MB/s.<br>
Kicking off a new read after the previous one had succeeded went straight to the second server,<br>with throughput still at ~25MB/s.<br><br>Is this the expected behavior for clients that have failed over from one server to another ?<br>
<br>torbjorn@srv18:~$ sudo gluster volume$<br>Volume Name: testvol$ <br>Type: Distributed-Replicate$ <br>Volume ID: 90636b5d-0d57-483c-bbfd-c0cdab2adaaa$<br>Status: Started$ <br>
Number of Bricks: 2 x 2 = 4$ <br>Transport-type: tcp$ <br>Bricks:$ <br>Brick1: srv18.trollweb.net:/srv/gluster/brick0$<br>Brick2: srv17.trollweb.net:/srv/gluster/brick0$<br>
Brick3: srv18.trollweb.net:/srv/gluster/brick1$<br>Brick4: srv17.trollweb.net:/srv/gluster/brick1$<br><br><br>--<br><div>Vennlig hilsen<br>Torbjørn Thorsen<br><br>Trollweb Solutions AS</div>