Date: Wed, 29 May 2019 13:56:20 -0600 From: Scott Long <scottl@samsco.org> To: Wallace Barrow <incin@incin.me> Cc: freebsd-scsi@freebsd.org Subject: Re: SAS3008 On FreeBSD 12 wGeli - OS Wont Boot Message-ID: <F19D332A-2E25-4D5F-9C82-220BD4EE6CD0@samsco.org> In-Reply-To: <41a1b364-4a17-44ff-a3ca-da9b80b16c2e@www.fastmail.com> References: <a62c7ed3-f5f1-4a5b-83df-949a32b0d8e8@www.fastmail.com> <120DF864-06FB-4396-887B-78DA3E7DA351@samsco.org> <3A9F7544-40A8-4446-ADC8-18CE57A9F55A@samsco.org> <14e0991e-2e53-4127-baf0-dc674b6d8653@www.fastmail.com> <f1aa37cb-9c0f-4b5e-aa69-05307da3345d@www.fastmail.com> <F80BA500-91BB-46D5-B630-524BA0A53364@samsco.org> <41a1b364-4a17-44ff-a3ca-da9b80b16c2e@www.fastmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Thanks! It looks like there are 2 problems here. One is that the chain = frame logic really needs to auto-tune itself and not complain on the = console (and have larger default scaling). The second is that under = extreme resource shortage caused by the lack of chain frames, the driver = has a bug that corrupts its internal state and bookkeeping. It=E2=80=99s = not clear to me why GELI would trigger that, but it=E2=80=99s a good = clue. Thanks again for the report, I=E2=80=99ll see if I can find root = cause. Scott > On May 29, 2019, at 1:53 PM, Wallace Barrow <incin@incin.me> wrote: >=20 > That is correct.=20 >=20 > After the tunable was set the servers were rebooted many times and I = am unable to reproduce the issue. So it seems that was the fix. >=20 > --=20 > Wallace Barrow > incin@incin.me >=20 > On Wed, May 29, 2019, at 2:49 PM, Scott Long wrote: >> Hi Wallace, >>=20 >> Sorry for the long delay in getting back to you, thanks for the = update.=20 >> Is my understanding correct that the problem went away once you=20 >> manually set the chain_frames tunable during install? >>=20 >> Thanks, >> Scott >>=20
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?F19D332A-2E25-4D5F-9C82-220BD4EE6CD0>