https://bugzilla.novell.com/show_bug.cgi?id=726635 https://bugzilla.novell.com/show_bug.cgi?id=726635#c0 Summary: [NetApp CQ190205] Mellanox Infiniband hosts reboot during straight I/O with 512 paths Classification: openSUSE Product: openSUSE 11.2 Version: Final Platform: x86-64 OS/Version: SLES 11 Status: NEW Severity: Normal Priority: P5 - None Component: Kernel AssignedTo: kernel-maintainers@forge.provo.novell.com ReportedBy: sean.stewart@netapp.com QAContact: qa@suse.de Found By: --- Blocker: --- User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/535.1 (KHTML, like Gecko) Chrome/14.0.835.202 Safari/535.1 Two hosts with SLES 11 SP 1, and Mellanox Connect-X QDR HCAs ran I/O to 128 volumes on a NetApp storage array, through a switch. Each volume had four paths from the host to the array, so a total of 512 paths were seen on the system. After running I/O to these 128 volumes on each system, one host rebooted after two minutes, and the other rebooted after 39 minutes. OS: SLES11 SP1 HBA: Mellanox Connect-X InfiniBand QDR Driver: Mellanox OFED 1.5.2 Reproducible: Didn't try Steps to Reproduce: 1. Map 128 LUNs to the host, so that four paths are seen by the host for each LUN 2. Start I/O 3. Actual Results: The host reboots, and generally slows down significantly. From the output of free, it does not look like a memory usage issue, but the host still becomes mostly unresponsive, possibly leading to the reboot. Expected Results: The system should stay up. Logs are difficult to capture for this problem, since the messages file shows absolutely no errors, and suddenly the system reboots. No kernel panic or crash dump is seen either. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.