Hi Chip, Chip Coldwell wrote: > On Thu, 13 Mar 2008, lemons_terry@xxxxxxx wrote: >> Sorry, but doesn't scsi_id only return what it finds in SCSI VPD pages >> 80 or 83? Not all devices provide a WWN in VPD page 83. > > It's quite a bit more complicated than that. If a WWN exists on VPD > pages 80 or 83, scsi_id will return that; otherwise it will synthesize > a serial number. > > Doug Gilbert and I have exchange some correspondence on the subject of > unifying his lsscsi (and possibly sg3_utils), my scsitool and the udev > scsi_id utility. At the moment, all of these are drawn from different > code bases. Doug chose to design lsscsi to use only sysfs information > because he wanted it to be runnable as a non-privileged process. > That's both good and bad: sysfs is shifting sand to build a foundation > on, and so far there is no standard place in sysfs to retrieve a WWN. > Both scsi_id and sg3_utils use SG_IO to do a VPD 0x80/0x83 INQUIRY. > I'd be more than happy to move the scsi_id program out of the udev codebase and merge it with sg3_utils. Or, possibly better, would be to have scsi_id use the libraries provided by sg3_utils. scsi_id is just part of udev as it accidentally ended up there ... IIRC we even hijacked that out of multipath-tools at one point. And what's more, it's effectively orphaned as Patrick Mansfield doesn't seem to work on SCSI things anymore. So moving scsi_id to a better maintained base would indeed be a good idea. BTW: what is the main reason for scsitool? Does it have features which scsi_id and sg3_utils are missing? Cheers, Hannes -- Dr. Hannes Reinecke zSeries & Storage hare@xxxxxxx +49 911 74053 688 SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg GF: Markus Rex, HRB 16746 (AG Nürnberg) -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
