Hello to all- Is there an easy way to tell if I can use a particular drive for burning a CD? Basically, I'm trying to set up a web front end for CD Recording here at work. We have a machine with 4 writers on it, and I want to allow a user to submit jobs to be recorded into the queue. Then as the drives become available, it will start recording. So I need to know if the drive is in use, and if it has media. I also want it to be robust in case someone comes along and starts their own burn outside of my program. I suppose I could just do a ps | grep cdrecord to determine which drives are currently in use. However, I still need to know if there is valid media (perhaps the last person when they went to pick up their CD didn't put it back in the drive) I also know that if a burn is in progress (not in the wait time, but actually in the burning stage) cdrecord -checkdrive will report that there are no supported modes. I know if I do cdrecord -dummy -v ..., it will not burn, but will attempt and will try to detect media, etc. However, this waits for 9 seconds to do an test burn before exiting. I suppose I could send a ^C into the stdin, it just in general seems like a clumsy way to do it. I realize this is a little unorthodox of a situtation, but if I get it working, it could be pretty useable. The basic flow for those who are interested... A web front end (written in python) that allows users to submit new entries into the queue, browse the queue, remove entries from it, etc. The front end requires the user to log in as well as supply an email address so they can get notifications. Then a cron job runs every minute and checks to see if there are any entries that haven't been burned yet. It then checks the list of drives for the first available one. If it doesn't find any, it will leave the item in the queue and wait. If an item has been in the queue too long, it will send an email back to the user saying it hasn't been able to fulfill the request yet. If there is an available drive, it will create 2 new processes. mkisofs and cdrecord. mkisofs will be run as the user (to prevent the user from burning to disk things that they shouldn't be able to see), but cdrecord has to run as root (at the very least setuid root, but might as well stay root) I'm not sure if I'll need to finish mkisofs before starting cdrecord. That will depend on network latencies, but I'm thinking to just give cdrecord a fairly large fifo buffer (like 10-20MB) and see if that works. We use a fairly underused 100Mb network, so I think it will be able to handle it (although with 4 burns going at the same time, I'll have to play around a little.)
That reminds me, I'm also getting some weird results when trying to burn two CDs at the same time. Basically if the drives are on the same IDE channel it gets REALLY flaky. In general it seems that when one of the burns finishes it will cause a 'scsi' error on the other drive. Any suggestions on how to fix this? I already have 1 add-in IDE controller so I have 4 available IDE channels (the hard-drive is on 1, so that leaves 3 free). And I'd really like to be able to use all 4 drives at the same time. Especially since eventually we'll move to DVDs which means much more time waiting for finished disks. Thanks for any help, John =:-> PS> I would consider releasing this if there is any interest. It's really only going to be a collection of about 10 python scripts. So I don't know if it's a big enough project to be worthwile.
smime.p7s
Description: S/MIME Cryptographic Signature
