Description of problem: Upgraded to the latest (2.6.18-1.2257), mounted the EMC filesystem via this entry in /etc/fstab: //filesrv1/news /mnt/win/filesrv1/news cifs workgroup=workgroup,credentials=/ usr/share/credentials,rw,uid=500,auto 0 0 Accessing the contents of any given file (say, via cat or less or some such) yields these in dmesg: [801027.110334] CIFS VFS: RFC1001 size 135 bigger than SMB for Mid=18 [801027.110344] Bad SMB: : dump of 48 bytes of data at 0xffff81003fb75700 [801027.110358] 00000087 424d53ff 000000a2 c0018000 . . . . S M B . . . . . . [801027.110376] 00000000 00000000 00000000 4035003f . . . . . . . . . . . . ? . 5 @ [801027.110395] 0012003f 0000ff2a 003f0100 00000001 ? . . . * . . . . ? . . . . . [801083.993639] CIFS VFS: server not responding [801083.993647] CIFS VFS: No response for cmd 162 mid 18 And these in /var/log/messages: Jan 4 16:41:56 quant16 kernel: [801027.110334] CIFS VFS: RFC1001 size 135 bigger than SMB for Mid=18 Jan 4 16:42:52 quant16 kernel: [801083.993639] CIFS VFS: server not responding Jan 4 16:42:52 quant16 kernel: [801083.993647] CIFS VFS: No response for cmd 162 mid 18 Version-Release number of selected component (if applicable): 2.6.18-1.2257 How reproducible: Easily Steps to Reproduce: 1. Upgrade kernel to 2.6.18.x 2. CIFS-mount EMC's CELLERRA array (suspect that other EMC products are similarly affected) 3. Cat a file on the mount point Actual results: Command hangs/times out. SIGKILL is honored. Expected results: Get contents of file. Additional info: Replicated on FC6 2.6.18-1.2489 x86_64
This is a known bug in EMC Celerras. There is more info in the RHEL5 equivalent case, bug #246093. To resolve this issue, upgrade to NAS Code 5.5.27.5 or later (use EMC Primus case number emc165978).