Hi, > Yes, in every case of blank disc insert there are five such messages: > [942204.231274] sr 59:0:0:0: [sr1] CDB: Read(10): 28 00 00 00 00 00 00 00 01 00 00 00
Ouchers. At least for CD and DVD- this is evil. DVD-RAM, DVD+, BD-RE are not affected. (And i though BD-R was safe, too.) > I will try this next time. They have nine lives and then they pop up as undead somnambules. > It is so strange that this problem only now > begins, and occurs 3 times in a row with xorriso > but not growisofs. If it is the cause, then it is a race condition. Small differences in the race start can cause a totally different outcome. Another theory could be about the SRM+POW formatting of growisofs. Maybe it is less prone to disturbance on your drive. You could format your blank BD-Rs by help of dvd+rw-format and then use them by cdrskin or xorriso. At the time of the failure, there has been few difference in the actions of growisofs and libburn. The use of POW by growisofs comes at the very end. > maybe they are still bothering > the drive during the burn. Try to disable everything that complains about the blank media. Only if those complaints are all gone, try whether cdrskin and xorriso are safe with their self-formatted or unformatted BD-R. On the other hand - given the price of BD-R - it seems more economic to wait whether growisofs fails to reach its final error in the next dozen tries. You could test cdrskin or xirriso with CD-RW media. They are very vulnerable to disturbance and can be blanked quickly after a coaster. > Is there no kernel mechanism to quiesce the drive and then lock it for > exclusive access? LKML says its up to userspace to coordinate. Userspace has no means for reliable mandatory locking of the physical device. > Is POW only needed for modifying UDF discs? growisofs uses it to copy the "superblock" of the newest ISO 9660 session over the superblock of the first session. So both, a smart operating system or a dumb OS, will mount the newest directory tree. I am not sure whether it happens with the first session at all. But dvd+rw-tools formats the media so that it can be done. > This is getting expensive! Thanks for all your help. My comiserations. I have few BD-R media and refrain from making own stress tests. I run two BD-Rs currently for daily incremental backups. No disturbers - no failures. (Although my first BD-R refused after 330 sessions to take another one. The drive said "write error". I decided to keep it readable rather than to try fixing its writeability.) Again from economic view: If BD-RE are not more than twice as expensive than BD-R, then forget the BD-R. In the sum you are much better off with BD-RE. Technically i am very interested in finding out what fails here. Have a nice day :) Thomas -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: http://lists.debian.org/[email protected]
