On Mon, 29 Jan 2018 15:04:45 -0600, John McKown wrote: >On Mon, Jan 29, 2018 at 2:41 PM, Bill Ashton wrote: >> >> He ran into an error using BPXBATCH for a PAX command, but could not find >> any documentation after searching on Google for nearly an hour. The error >> message looks like this - *pax: checksum error on tape (got #####, expected >> 0)* > >I think this means that the PAX archive is corrupt. Did you ensure that >it's trail amongst all systems was a BINARY one? > And use cksum(1) at all steps along that trail.
>I am not an expert on this, but I think that there is a "hash value" >(checksum) associated with the PAX archive. The PAX command which creates >the archive adds in some "extra" so that the overall checksum is zero. So, > That can be done only for a fairly weak checksum, else bitcoin would be very cheap. >IOW, the #### value does not have any real meaning, ... > >ref: http://pubs.opengroup.org/onlinepubs/009695299/utilities/pax.html >=== >... >> and the ##### is some hex value, like eb0a or 1015b. He got this even >> though he wasn't even using tape - must be a standard message. >> They oughta fix that. The Command Ref. uses "archive file"; the message should agree. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
