Date: Sun, 24 Mar 2002 15:09:22 -0800 From: Michael Smith <msmith@freebsd.org> To: F J <fj6712@yahoo.com> Cc: freebsd-scsi@FreeBSD.org Subject: Re: SCSI write semantics Message-ID: <200203242309.g2ON9MO02559@mass.dis.org> In-Reply-To: Your message of "Sat, 23 Mar 2002 10:54:30 PST." <20020323185430.80848.qmail@web20503.mail.yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> Are the SCSI disk writes atomic or they're hardware > dependent? They are not atomic. > What I mean is that if you issue a > 32KB write and the power goes off at the same > time, is there any gauarantee that either the > disk blocks will be updated completely or > nothing will be written at all? No guarantees. > what about SCSI RAID > controllers? Controllers with battery-backed memory will typically attempt to complete= = writes that were outstanding when power is reapplied and the array = verified to match the pre-power-loss configuration. This is about the = only configuration that gives you any "guarantees" at all. > Another concern is that DMA may be in > progress while memory contents become unstable. Is > this possible? I wouldn't expect this to be a problem; by the time memory has become = unstable, the rest of the system will be dead (and I suspect that the = chipset would be suppressed as soon as soon as the rails go out of spec).= -- = To announce that there must be no criticism of the president, or that we are to stand by the president, right or wrong, is not only unpatriotic and servile, but is morally treasonable to = the American public. - Theodore Roosevelt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200203242309.g2ON9MO02559>