Open enealDC opened 10 years ago
We've hit the same crash two times (for now) on 3.16.2-201.fc20.x86_64.
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.512546] ------------[ cut here ]------------
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.512632] kernel BUG at /root/devel/EnhanceIO/Driver/enhanceio/eio_main.c:1522!
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.512775] invalid opcode: 0000 [#1] SMP
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.512853] Modules linked in: vhost_net vhost macvtap macvlan tun binfmt_misc drbd lru_cache ip6table_filter
ip6_tables ebtable_nat ebtables bridge stp llc enhanceio_rand(OE) enhanceio_lru(OE) enhanceio_fifo(OE) enhanceio(OE) iTCO_wdt gpio_ich iTCO_vendor_su
pport ppdev raid1 lpc_ich mfd_core coretemp kvm_intel kvm microcode snd_intel8x0 snd_ac97_codec serio_raw i2c_i801 ac97_bus i2c_core snd_seq snd_seq_d
evice snd_pcm snd_timer snd soundcore ixgbe e1000e ptp pps_core mdio i5000_edac parport_pc edac_core parport dm_thin_pool dm_persistent_data dm_bio_pr
ison libcrc32c i5k_amb ioatdma dca shpchp nfsd auth_rpcgss nfs_acl lockd sunrpc ata_generic pata_acpi arcmsr
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.513231] CPU: 1 PID: 28242 Comm: kworker/u8:0 Tainted: G OE 3.16.2-201.fc20.x86_64 #1
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.513231] Hardware name: Supermicro X7DA8/X7DA8, BIOS 6.00 10/02/2006
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.513231] Workqueue: eio_callback eio_post_io_callback [enhanceio]
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.513231] task: ffff880212378000 ti: ffff8801f0f10000 task.ti: ffff8801f0f10000
Nov 21 10:08:05 10.0.12.177 kernel: [1475290.513231] RIP: 0010:[
Hello. My hardware details are as follows:
Power Edge R515 AMD Procs Fusion-IO Duo 640gb (latest firmware and drivers) (cache) Dell H700 RAID5 (12 15K 600GB SAS) (backing)
I'm using SCST to expose the cache backed device via SRPT to a ESX5.5 Initiator. I've mapped that LUN to an RDM disk which I'm using an ASM disk for an Oracle 11gR2.
I'm using the SLOB benchmark tool and when I'm just seeding the tables, I'm presuming after the cache gets warmed up and etc, I get a bad crash.
I can consistently reproduce this so I setup netconsole to log it to a remove machine.
http://pastebin.com/vMMpzTkS
My goal is to examine bcache, dmcache, enhancio and zfsonlinux to see if they really deliver the promise of hybrid storage in an IAAS scenario.
Any thoughts you could share would be useful