e i g h t - c u b e d . c o m 

Fast I/O problems again

Quite some time ago, I described a problem where system service $IO_PERFORMW would return an unexpected status of SS$_FANDLEBUSY.

I reported this problem to Engineering and they fixed it some time later.

Well, yesterday, I had a report from a DBM user, who saw this problem. When I suggested that a patch was needed, I discovered the user was running with a patch later than the fixed version.

Have Engineering dropped a fix from a patch stream again?

When the original patch came out fixing this problem, we had just done a patch round on all our machines, and by the time that we got to doing another patch round, VMS732_SYS-V0500 had been released. So we applied that and I confirmed that the problem was fixed.

The DBM user reporting the problem yesterday indicated that they were running with VMS732_SYS-V0600 on, and as I'm running with V0700 on, I did a test run on the code, and sure enough, it failed. (However, then I tried to repeat that failure and was unable to).

I reported the problem to HP support, so stay tuned for updates.