I was not sure if misc or bugs would be appropriate for this problem, so it's on misc now.
As described in the title, I'm experiencing kernel panics with OpenBSD 6.0 running in VMWare ESX when a SAN Failover happens. The problem is descibed here: https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1009465 and here https://www.dan.me.uk/blog/2009/05/24/freebsd-with-esx/ http://nex7.blogspot.co.at/2013/03/vm-disk-timeouts.html The Failover on our SAN only drops 1 ping when we tested it last time and all other machines except OpenBSD boxes can handle the disk timeout. I'm looking for a way to increase SCSI timeouts in OpenBSD, does anyone know of a way i can solve this probem? Looking through all available sysctl parameters didn't get me anywhere. Heres a log which gets printed on console when such a failover happens (look like it panics when trying to allocate an inode): mode = 0100660, inum = 8, fs = /tmp panic: ffs_valloc: dup alloc Stopped at Debugger+0x9: leave TID PID UID PRFLAGS PFLAGS CPU COMMAND *69032 48643 502 0x13 0x4000000 3 mysqld 78949 78949 73 0x100010 0x80 2 syslogd 97908 97908 0 0x14000 0x210 1 softnet Debugger() at Debugger+0xfe panic() at panic+0xfe ffs_inode_alloc() at ffs_inode_alloc+0x408 ufs_makeinode() at ifs_makeinode+0x85 ufs_create() at ufs_create+0x34 VOP_CREATE() at VOP_CREATE+0x3d vn_open() at vn_open+0x2cd doopenat() at doopenat+0x187 syscall() at syscall+0x27b -- syscall (number 5) -- end of kernel end of trace frame: 0xe99b406dd10, count: 6 0xe9945c0580a: I missed creating a ddb trace and ddb show uvm, I can provide further info should a SAN failover happen again (hopefully not). Thanks in advance for any help. Mario
