Here are some recent things I have done, as promised:
1. I backed up the monitor scratch area and deleted some of the
old stuff.
2. I talked to Marini yesterday. He said that he increased the disk
quota on MICHPUB by some 10's of thousands of blocks (I forget the
exact number) for the purpose of having space for monitor output.
This will solve the problem of monitor crashes when the scratch disk
fills up. However, it also means that MICHPUB users must not fill up
this extra disk space. So... if you use the MICHPUB area, PLEASE
always leave *at least* 10000 blocks free. I will personally make
sure that the monitor output sitting around on disk never exceeds
40000 blocks (several months' worth).
I made the new monitor output directory [MICHPUB.ERP.ERPGC.OUT].
3. I also asked Marini about the remail attempt interval for SMTP on
VXMACB. He said he didn't know what it was, but he would look into
it.
4. For better redundancy for email to the beeper, the beeper mail
message gets sent twice now, via two different paths out of the tunnel
(by SMTP directly, and also by DECNET to VAXGS first). This is in
hope of increasing email speed when the tunnel-to-outside link gets
flaky (although it's not clear whether DECNET mail will work faster if
Internet mail gets delayed -- time will tell.) Unfortunately it means
that the beeper will beep twice for the same alarm.
5. As some of you already know, I updated Calmod on VXMACB, so that
now the most recent calibration constants are available to the monitor
for event reconstruction.
Kate.