Date: Tue, 26 May 2015 09:16:23 -0700 From: Alexander Motin <mav@FreeBSD.org> To: Neel Natu <neelnatu@gmail.com>, Garrett Cooper <yaneurabeya@gmail.com> Cc: FreeBSD Current <freebsd-current@freebsd.org>, "freebsd-testing@freebsd.org" <freebsd-testing@freebsd.org>, Peter Grehan <grehan@freebsd.org> Subject: Re: cam(4) timeouts in bhyve/kyua runs up on Jenkins? Message-ID: <55649C57.1080906@FreeBSD.org> In-Reply-To: <CAFgRE9HR5AOeTqOfFD4Wzg3Zbp%2BF4n1_MOBzd8k%2B1qJjVhp_AA@mail.gmail.com> References: <28845361-E624-4CB9-B255-1B2C1FD0DF28@gmail.com> <4BEE0A54-3AD2-42DB-AF4E-2B0A5E34DF4E@gmail.com> <553F3CA7.9090601@FreeBSD.org> <9FBE053A-971A-433E-9B88-D62CF9273CFB@gmail.com> <CAFgRE9HR5AOeTqOfFD4Wzg3Zbp%2BF4n1_MOBzd8k%2B1qJjVhp_AA@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 26.05.2015 09:06, Neel Natu wrote: > On Mon, May 25, 2015 at 1:40 PM, Garrett Cooper <yaneurabeya@gmail.com> wrote: >> On Apr 28, 2015, at 0:54, Alexander Motin <mav@FreeBSD.org> wrote: >>> On 27.04.2015 21:17, Garrett Cooper wrote: >>>> On Apr 27, 2015, at 11:16, Garrett Cooper <yaneurabeya@gmail.com> >>>> wrote: >>>> >>>>> I was looking at the console log for the latest kyua run and I’ve >>>>> noticed that it’s timing out a bit more [1] than it was >>>>> previously [2]. I’ve seen some of your commits recently to cam(4) >>>>> dealing with bhyve — has there been a performance regression >>>>> there? Thanks! -NGie >>>>> >>>>> 1. >>>>> https://jenkins.freebsd.org/job/FreeBSD_HEAD-tests2/940/console >>>>> 2. >>>>> https://jenkins.freebsd.org/job/FreeBSD_HEAD-tests2/983/console >>>> >>>> (Sorry for not being more explicit for the archives) These are the >>>> timeouts I’m referring to: >>>> >>>> ahcich0: is 00000000 cs 00000000 ss 1f000000 rs 1f000000 tfd 50 >>>> serr 00000000 cmd 1000dc17 (ada0:ahcich0:0:0:0): >>>> WRITE_FPDMA_QUEUED. ACB: 61 08 a8 54 1e 40 00 00 00 00 00 00 >>>> (ada0:ahcich0:0:0:0): CAM status: Command timeout >>>> (ada0:ahcich0:0:0:0): Retrying command > > Can you try to reproduce with a virtio-blk instead of ahci-hd? > > It will help narrow down the problem because both ahci-hd and > virtio-blk emulations share a common backend to read/write to the > disk. virtio-blk has no command timeouts, so it may be problematic to notice delayed commands there, unless they stuck forever. This won't be fair comparison. Though if not use TRIM, virtio-blk should be more efficient now, as it is supposed to. >>> Last time I was more working on bhyve host disk emulation, rather then >>> on cam(4) running on guest. Considering that, what guest and what host >>> versions are you running? Is there any other load on host except this VM >>> that could cause I/O delays high enough to trigger timeouts? What are >>> you using to back the virtual disk (file, zvol, ...)? >> >> I have no idea what the Jenkins slaves are running in terms of configuration/version/etc. You’ll have to ask jenkins-admin@… >> Thanks! >> -NGie -- Alexander Motin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?55649C57.1080906>