Re: BD-RE DL write failures
Hi, > I suspect the good/slow items are again the media. I assume that most of the "slow" ranges are due to relocated blocks. Defect Management decided that the read quality was not good enough and put a copy into the Spare Area, which is located at the inner rim of the medium. So the laser head has to hope inwards and then back outwards when it reaches the original invalidated position of the chunk. > At the present time I am of the opinion that BD-RE is just too > unreliable to continue working with it for data purposes. The combination of BD-RE DL and your drives does not work well, indeed. My own experience with single layer BD-RE is different. Not more than two or three die per year. And i am using them daily. > I might try another method of using the media like a tape drive by > writing the tar file directly. I think xorriso will do that from the > man page. It can put an ISO 9660 filesystem onto a tape, but that would be hard to read. In the old days i put tar or afio archives directly on QIC tapes. It is quite painful to wait minutes until a particular file was found and restored. Consider to use a more flexible tape archiver like "dar". In any case it is important that you practice restoring of particular files before you decide for an archive format. Have a nice day :) Thomas
Re: BD-RE DL write failures
I set stream_recording off yesterday and did a run at the end of the day. The screen output appeared that it worked OK. Then the cron ran the script this morning and it took over 2 hours. But it failed again, however in different locations. xorriso : UPDATE : 2375392 blocks read in 951 seconds , 1.5xB xorriso : UPDATE : 2378752 blocks read in 952 seconds , 1.5xB libburn : SORRY : SCSI error on read_10(2379264,32): [3 11 05] Medium error. L-EC uncorrectable error. xorriso : UPDATE : 2379296 blocks read in 957 seconds , 0.0xB xorriso : UPDATE : 2382496 blocks read in 958 seconds , 1.5xB and xorriso : UPDATE : 9321120 blocks read in 2559 seconds , 2.1xB xorriso : UPDATE : 9325696 blocks read in 2560 seconds , 2.1xB libburn : SORRY : SCSI error on read_10(9329312,32): [3 11 05] Medium error. L-EC uncorrectable error. xorriso : UPDATE : 9329344 blocks read in 2563 seconds , 0.6xB xorriso : UPDATE : 9331904 blocks read in 2564 seconds , 1.2xB and xorriso : UPDATE : 10593k blocks read in 2897 seconds , 2.1xB xorriso : UPDATE : 10598k blocks read in 2898 seconds , 2.1xB libburn : SORRY : SCSI error on read_10(10856736,32): [3 11 05] Medium error. L-EC uncorrectable error. xorriso : UPDATE : 10602k blocks read in 2901 seconds , 0.7xB xorriso : UPDATE : 10604k blocks read in 2902 seconds , 1.0xB This is exactly the same media and drive as earlier. The problems in different locations. That has been typical of all the media as well. I have not placed any stickers or labels on the media that can cause balance issues. The final results of check media below. Sorry about the output being so long. I suspect the good/slow items are again the media. It reports 4 as unreadable so I must have missed one. xorriso : UPDATE : 12362k blocks read in 3523 seconds , 2.1xB xorriso : UPDATE : 12363k blocks read in 3523 seconds = 1.6xB Media checks :lba , size , quality Media region : 0 , 1760 , + good Media region : 1760 , 32 , + slow Media region : 1792 ,2377472 , + good Media region :2379264 , 32 , - unreadable Media region :2379296 ,1014944 , + good Media region :3394240 , 32 , - unreadable Media region :3394272 , 32 , + slow Media region :3394304 ,960 , + good Media region :3395264 , 32 , + slow Media region :3395296 ,960 , + good Media region :3396256 , 32 , + slow Media region :3396288 ,960 , + good Media region :3397248 , 32 , + slow Media region :3397280 ,960 , + good Media region :3398240 , 32 , + slow Media region :3398272 ,960 , + good Media region :3399232 , 32 , + slow Media region :3399264 ,5930048 , + good Media region :9329312 , 32 , - unreadable Media region :9329344 ,1527392 , + good Media region : 10856736 , 32 , - unreadable Media region : 10856768 , 65440 , + good Media region : 10922208 , 32 , + slow Media region : 10922240 , 22368 , + good Media region : 10944608 , 32 , + slow Media region : 10944640 , 79072 , + good Media region : 11023712 , 32 , + slow Media region : 11023744 , 12384 , + good Media region : 11036128 , 32 , + slow Media region : 11036160 , 1568 , + good Media region : 11037728 , 32 , + slow Media region : 11037760 , 9984 , + good Media region : 11047744 , 32 , + slow Media region : 11047776 , 60928 , + good Media region : 11108704 , 32 , + slow Media region : 11108736 , 7584 , + good Media region : 6320 , 32 , + slow Media region : 6352 , 6560 , + good Media region : 11122912 , 32 , + slow Media region : 11122944 , 5952 , + good Media region : 11128896 , 32 , + slow Media region : 11128928 , 18432 , + good Media region : 11147360 , 32 , + slow Media region : 11147392 , 13120 , + good Media region : 11160512 , 32 , + slow Media region : 11160544 , 2496 , + good Media region : 11163040 , 32 , + slow Media region : 11163072 ,896 , + good Media region : 11163968 , 32 , + slow Media region : 11164000 , 1024 , + good Media region : 11165024 , 32 , + slow Media region : 11165056 , 23168 , + good Media region : 11188224 , 32 , + slow Media region : 11188256 , 6592 , + good Media region : 11194848 , 32 , + slow Media region : 11194880 , 1536 , + good Media region : 11196416 , 32 , + slow Media region : 11196448 , 635712 , + good Media region : 11832160 , 32 , + slow Media region : 11832192 , 73952 , + good Media region : 11906144 , 32 , + slow Media region : 11906176 , 2048 , + good Media region : 11908224 , 32 , + slow Media region : 11908256 ,352 ,
Re: BD-RE DL write failures
Hi, > libburn : SORRY : SCSI error on read_10(297920,32): [3 11 05] Medium error. > L-EC uncorrectable error. That's bad news for your hardware. A data chunk of 64 KiB did not match its inner checksums. So the drive refuses to hand out these 64 KiB. You would have to try the media in other drives to get an impression whether it's the media or your drive which are to blame. > Media checks :lba , size , quality > Media region : 297824 , 32 , - unreadable > Media region : 297920 , 32 , - unreadable > Media region : 298048 , 32 , - unreadable Just three unreadable chunks, quite near together. Not even 600 MiB away from the medium start. If it was necessary you could copy the damaged image to a data file on hard disk (e.g. $HOME/bdre.img): xorriso -indev /dev/sr0 -check_media data_to="$HOME"/bdre.img -- You could even try several times or on several BD drives by having a file bdre.map which records which blocks were read successfully and which are yet missing in bdre.img: xorriso \ -indev /dev/sr0 \ -check_media data_to="$HOME"/bdre.img sector_map="$HOME"/bdre.map -- The already successfully read blocks will not be tried again. One may carry medium, bdre.img, and bdre.map from computer to computer in order to find a drive which can read the missing blocks. The fact that command -indev did work shows that the directory tree is not hit by the bad spots. So, perfectly restored or not, the file bdre.img should be mountable and the damaged tar archive should be readable. Dunno what tar will do when it encounters 64 KiB of zeros in the middle of an archive where it would expect some of its metadata. This is a good opportunity to practice for a case of media which are ok after recording but turn out to be slightly rotten much later. But well, this is also the classic situation for Defect Management: Some few bad spots on a large, elsewise good medium. So you should try with one of the problematic media whether it works slowly but correctly if you set -stream_recording "off" It might be that Defect Management goes mad and lets the drive gnaw endlessly not only on the bad spots above, but also on many other spots. Or it might be that you have a rare case where Defect Management fulfills the hopes which it stirs up. Have a nice day :) Thomas
Re: BD-RE DL write failures
On Wed, 2017-08-16 at 19:34 +0200, Thomas Schmitt wrote: > Hi, > > sorry, i did not react properly on this piece of info: > > > > > The format, erase, and burn processes all complete normally with no > > issues or errors. When I mount and attempt to read the tar file it > > aborts with an error message. > Do you see related messages in the system logs ? > > What happens if you let xorriso read the medium: > > xorriso -indev /dev/sr0 -check_media -- > > > Have a nice day :) > > Thomas > The last of the output from the above command on a usable media. xorriso : UPDATE : 12276k blocks read in 1291 seconds , 6.0xB xorriso : UPDATE : 12289k blocks read in 1292 seconds , 6.0xB xorriso : UPDATE : 12294k blocks read in 1292 seconds = 4.4xB Media checks :lba , size , quality Media region : 0 ,160 , + good Media region :160 , 32 , + slow Media region :192 ,128 , + good Media region :320 , 32 , + slow Media region :352 , 32 , + good Media region :384 , 64 , + slow Media region :448 , 12588945 , + good Media region : 12589393 , 11062959 , 0 untested On media that is not usable, I get the following error when I try to use tar and read the file. gzip: stdin: Input/output error tar: Unexpected EOF in archive tar: Error is not recoverable: exiting now I attempted to run the same command above on the failed burn and I received the following during the process. xorriso : UPDATE : 292736 blocks read in 70 seconds , 1.4xB xorriso : UPDATE : 296448 blocks read in 71 seconds , 1.6xB xorriso : UPDATE : 297824 blocks read in 72 seconds , 0.5xB libburn : SORRY : SCSI error on read_10(297824,32): [3 11 05] Medium error. L-EC uncorrectable error. xorriso : UPDATE : 297856 blocks read in 76 seconds , 0.0xB xorriso : UPDATE : 297888 blocks read in 78 seconds , 0.0xB libburn : SORRY : SCSI error on read_10(297920,32): [3 11 05] Medium error. L-EC uncorrectable error. xorriso : UPDATE : 297952 blocks read in 82 seconds , 0.0xB xorriso : UPDATE : 298016 blocks read in 84 seconds , 0.0xB libburn : SORRY : SCSI error on read_10(298048,32): [3 11 05] Medium error. L-EC uncorrectable error. xorriso : UPDATE : 298080 blocks read in 88 seconds , 0.0xB xorriso : UPDATE : 300640 blocks read in 89 seconds , 1.2xB xorriso : UPDATE : 304576 blocks read in 90 seconds , 1.8xB xorriso : UPDATE : 308288 blocks read in 91 seconds , 1.7xB The program continued to run and then completed with the following. xorriso : UPDATE : 12337k blocks read in 2787 seconds , 2.1xB xorriso : UPDATE : 12342k blocks read in 2788 seconds , 2.1xB xorriso : UPDATE : 12342k blocks read in 2788 seconds = 2.1xB Media checks :lba , size , quality Media region : 0 , 264736 , + good Media region : 264736 , 32 , + slow Media region : 264768 , 33056 , + good Media region : 297824 , 32 , - unreadable Media region : 297856 , 32 , + slow Media region : 297888 , 32 , + good Media region : 297920 , 32 , - unreadable Media region : 297952 , 32 , + good Media region : 297984 , 32 , + slow Media region : 298016 , 32 , + good Media region : 298048 , 32 , - unreadable Media region : 298080 , 12340187 , + good Media region : 12638267 , 11014085 , 0 untested xorriso : NOTE : Tolerated problem event of severity 'SORRY' xorriso : NOTE : -return_with SORRY 32 triggered by problem severity SORRY The script is run by cron during the night. Reliability is more important than performance. So I have set stream_recording off for tonight on this media. I will see how it completes in the morning. Thank you, James
Re: BD-RE DL write failures
Hi, sorry, i did not react properly on this piece of info: > The format, erase, and burn processes all complete normally with no > issues or errors. When I mount and attempt to read the tar file it > aborts with an error message. Do you see related messages in the system logs ? What happens if you let xorriso read the medium: xorriso -indev /dev/sr0 -check_media -- Have a nice day :) Thomas
Re: BD-RE DL write failures
Hi, > The media was labeled "for Video". So I found some Maxell that was > labeled "for Data" If those "for Video" were more expensive than the others, then you probably paid royalties to the movie industry. We had such a situation with special CD-R media for audio recording. Some living room CD recorders did not accept the less expensive normal CD-R. In general, there is no difference in BD and DVD recording between video and data. It's just a matter of filesystem (video prescribes UDF), particular files, and possibly cryptographic information for content authentication. Other than with DVD+R DL, it is not possible to chose the layer jump address with multi-layer BD-R[E]. Afaik, there is few way to distinguish a single layer BD-RE from a dual layer BD-RE other than by their size. So i would not expect that the dual layer BD-RE need any extra treatment. > Two have worked very well every time > for my 26 GByte tar file. The other three discs have the same issue as > the earlier media. > ... > Any suggestion on how to get a reliable burn? What's the symptoms exactly ? What messages do you get from xorriso ? - Preliminary thoughts. They might become useless when more info arrives. > /usr/bin/xorriso -stream_recording on Did you try with -stream_recording "off" ? That will be darn slow. If it clonks and goes to 0.0x speed with only a 0.1x every fifth pacifier messages, then Spare Area blocks get employed. (Normally you get a medium error some time later.) > -speed "-1" That's equivalent to -speed "min". Just for the records. > I also added the -speed "-1" option to attempt to > lower the write speed but it did not have any effect BD-RE as overwritable media do not offer much choice of write speed and normally do not obey write speed settings, especially if you mark the data as urgent-to-write by -stream_recording "on". Have a nice day :) Thomas
