[Gluster-devel] A crash due to a invalid indoe table(glusterfs-3.3.1)

[deming jia]

Hi,All.

       I do a test in glusterfs-3.3.1,but have a crash，seems return an invalid indoe or invalid inode-table in stripe_lookup_cbk.

       The test likes this:

       1. craate a dht+stripe(3) volumes.

       2.mount in three different machines

       3.mkdir testdir in mount point

       4.every machine start a postmark to craate a lot of little files named 1,2,3..... in testdir

      

       In short，three postmark process in three different machine create a lot of little files named 1,2,3,4..... in a same dir （testdir）

use a dht+stripe volumes, this crash can be reproduce.

      I used xfs as the underlying filesystem.

     I used three postmark procresses on one client or  used three postmark processes on three clients of different mathines to fopen fwrite files named (client1:a1, a2, a3, a4....  

client2:b1, b2, b3, b4....  client3:c1, c2, c3, c4...) in the same dir , They all didn't produce a crash.

      Does creating a same file in the same dir  by different client of different matchines at the same time  have any problme?

      In additionally， http://supercolony.gluster.org/pipermail/gluster-devel/2011-August/007441.html , it's decription is similar to my 

crash. How can I solve it？ Look forward to your reply,Thanks.

       I have little English,sorry!

                                                   

                                                                                                                                                                                                                          Yours Jia Deming

                                                                                                                                                                                                                                2013-07-07