Date: Wed, 27 Sep 2000 18:25:05 -0500 From: David Kelly <dkelly@hiwaay.net> To: Terry Lambert <tlambert@primenet.com> Cc: dan@langille.org, chat@FreeBSD.ORG Subject: Re: CD writers - recommendations Message-ID: <200009272325.e8RNP5a28632@nospam.hiwaay.net> In-Reply-To: Message from Terry Lambert <tlambert@primenet.com> of "Wed, 27 Sep 2000 07:34:33 -0000." <200009270734.AAA20534@usr05.primenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Terry Lambert writes: > NB: The numbers on the Plextor are skewed down, since I have > not burnt 144 disks on the thing with a loss of 2, so I don't > know anything other than the statistical failure; on the other > hand, the 1 in 14 number for the IDE I used (don't know the > brand, sorry: it was supposed top of the line) is a valid > statistic over about 50 disks. Having done a couple hundred in prior employment, I'd say 1:72 is about right for *media* failure. Its not as if they can 100% test a CD-R. My prime burning machine was a Pentium 133, 24 MB, FreeBSD 2.2.5 to 2.2.8 (forgot), Adaptec 2940A, and most everything SCSI was external, one device per box, and probably over the limit on SCSI cable length. The CDR was a Yamaha CDR-100 or -102 (one of the first 4x). Kept images on Seagate (2) 4G Barracudas, the 1st generation, the STxxx150 number comes to mind. Had at least one each 4mm and 8mm tape drive. Stuff got moved around a fair bit. Had the vn device configured in the kernel so I could mount my images for verification. When duplicating CD's, would dd 2 blocks less than a cdrecord inquiry would report were on the disk to avoid the I/O error that would ensue if we tried to read them. Grab an md5 signature of the image, and a manifest of md5's of the files on the disk. Then I'd mount that image and recursive diff verify it against the original still mounted in the internal IDE CDROM. Often while burning it on the CD-R at the same time. Then I'd always put the new CD-R in the IDE drive, mount, and verify again against the md5 signatures. All of this verifying paid off, as others in our business were using Windows and were often shipping coasters. This was all low volume 1 to 5 disc runs. Boss really wanted me to do it in Windows too, but we never found a good way to verify the results. Decided the DJGPP Win32 Unix Utilities was overkill to get diff and md5sum. Made about 2 coasters for unknown reasons. More for having written a bad image in the first place (oops, it was a multi-session master). Got to the point if something needed to be done and there was a hole to put it in, did it, no matter what state the CD-R was in. Used tcopy a lot. A lot of the time the CD was copied from tape. Machine was also connected to the company network, ran Apache, samba, and Netatalk. XFree86. And was my primary desktop computer. The only problem was the vn code had a gotcha where the kernel might panic after umounting an image and then detaching it from the vn. Learned not to do that when the CD-R was burning. -- David Kelly N4HHE, dkelly@hiwaay.net ===================================================================== The human mind ordinarily operates at only ten percent of its capacity -- the rest is overhead for the operating system. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200009272325.e8RNP5a28632>