Date: Fri, 09 Dec 2016 15:56:58 +0100 From: Harry Schmalzbauer <freebsd@omnilan.de> To: Steven Hartland <killing@multiplay.co.uk> Cc: freebsd-scsi@freebsd.org Subject: Re: sporadic CAM (all devices) outage on 11-stable, mps(4), ahci(4) and bhyve(8) involved. Message-ID: <584AC63A.60301@omnilan.de> In-Reply-To: <a2348c42-bd76-245a-062a-d182bfc3113d@multiplay.co.uk> References: <584ABE6B.8060703@omnilan.de> <a2348c42-bd76-245a-062a-d182bfc3113d@multiplay.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
Bezüglich Steven Hartland's Nachricht vom 09.12.2016 15:43 (localtime): > Bad memory or CPU? Quiet sure that this can be ruled out. It's a IvyBridge Xeon and ECC memory and neither BMC nor the OS registrerd any MCE. And, the machine is still operational (from cache, true for both, guest and host) until something from disk is needed... The machine doesn't show any problems without heavy bhyve-ahci-load. I'll see if changing ahci, to virtio-blk changes anything. At least for this other still (r309605) present odditiy https://lists.freebsd.org/pipermail/freebsd-stable/2016-October/086157.html it doesn't make a difference, while I found virtio-blk beeing some orders of magnitude faster for unlinking UFS files - most likely due to lacking TRIM supprt. But on bare metal deleting a object tree on a SSD (with enabled TRIM support) doesn't take 10 minutes... Thanks, -Harry
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?584AC63A.60301>