| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 351.1 | | CAFEIN::PFAU | You can't get there from here | Wed Nov 05 1986 15:37 | 6 |
| The best I can remember around here (parvax, cafein) is about 35
days.
80 days? That's almost three months!!!
tom_p
|
| 351.2 | | SQM::RICO | | Thu Nov 06 1986 10:53 | 1 |
| What about PM? No PM in 80 days??
|
| 351.3 | PM ? We have PM every afternoon ! | MARVIN::WARWICK | Whack your porcupine | Thu Nov 06 1986 13:55 | 2 |
|
|
| 351.4 | 52 and counting | TOPCAT::GEISENHAINER | | Thu Nov 06 1986 14:05 | 3 |
| My uVAX (HULK) has been up for 52 days. It will most likely stay
up until we have a power hit or I get my hands on SDC V4.5... If
that works out to over 80, I'll post it here...
|
| 351.5 | A WHAT?? No WAY! | ASIA::MCLEMAN | Enjoy Oregon Wines! | Thu Nov 06 1986 14:07 | 7 |
| re: .2
PM's??? Do they still do those?? I haven't seen one on our systems
for a while. :-)
Jeff
|
| 351.6 | How about 108 days? | MDADMN::EATOND | Dan Eaton | Thu Nov 06 1986 14:11 | 7 |
| One of my field service buddies reports an 11/730 that ran for 108
days. The system was brought down then to upgrade VMS otherwise
it might of made it out to around 178 days since its on a 6 month
PM schedule.
Dan Eaton
|
| 351.7 | | ERIS::CALLAS | O jour frabbejais! Calleau! Callai! | Fri Nov 07 1986 10:34 | 7 |
| ERIS was up for 62 days once. Would have been longer if the *%$&*%
people building ZK3 hadn't cut the power lines to ZK1!
TRIFID (the VMS QAR system), in the days when it was a 730, was
once up for 91 days. And this was on FT1 of VMS V4!
Jon
|
| 351.8 | 95 days so far for a home system | FALEK::FALEK | ex-TU58 King | Fri Nov 07 1986 12:55 | 2 |
| Tim Garrison (a guy in our group)'s home microvax 2 has been up
for 95 days so far. His landlord pays the electric bill...
|
| 351.9 | PM = Promotional Maintenance | TURRIS::AMARTIN | Alan H. Martin | Sun Nov 09 1986 13:20 | 3 |
| The reliability of one of the KL's I used in Marlboro (KL2137? MRFORT?)
went up after we increased the time between PMs.
/AHM
|
| 351.10 | PM = Provocative Maintenance | SHEILA::PUCKETT | Open the pod bay doors please HAL | Mon Nov 10 1986 00:47 | 0 |
| 351.11 | PM = Premeditated Maintainance | PHENIX::SMITH | William P.N. (Wookie::) Smith | Mon Nov 10 1986 10:20 | 3 |
| That's usually performed by the system mangler, unless you have
a Eunichs system with a 'sysmgr', which is pronounced sis mugger.
|
| 351.12 | 18 months ................. | TROPPO::RICKARD | Doug Rickard - waterfall minder. | Tue Nov 11 1986 07:52 | 19 |
|
A customer in Sydney, Australia had a 780 that stayed up for
approx 18 months. Fixed applications, did not (WOULD NOT) upgrade VMS,
and restricted field service to doing voltage checks and seeing if the
fans were going at PMs. System finally came to grief because there is a
funny with the clock. At boot time, the time read from the battery
backup clock forms the high order part of the time-date, the low order
part is derived from some internal high speed interrupt. It turns out
that there is no carry from the low order to the high order, so that
after the system had been up for about 18 months without a reboot, ALL
SYSTEM TIMES WENT WRONG. This affected all file stamping, etc, etc, and
really caused a screw up. Engineering were finally able to tell us that
we had to reboot at least once every 18 months. By the way, it was about
VMS V1.6, so maybe thats saying something.......
Doug.
|
| 351.13 | | REGINA::OSMAN | and silos to fill before I feep, and silos to fill before I feep | Tue Nov 11 1986 15:43 | 10 |
| Oh well. So much for using VMS to manage a long-trip rocket voyage !
Unless of course the more recent VMS versions have been fixed to allow
more than 18 months of up-time.
Of course, the only for-sure way is to TEST it. Perhaps every new
release should be started on a machine in a corner somewhere and left
up FOREVER or until it crashes, as a test for longevity.
/Eric
|
| 351.14 | | ULTRA::PRIBORSKY | Tony Priborsky | Tue Nov 11 1986 15:59 | 3 |
| 18 months without some kind of power outage? Wow. DEC certainly
COULDN'T do it - we religiously take down entire plants once a year.
Must need to PM the transformers or something...
|
| 351.15 | | ERIS::CALLAS | O jour frabbejais! Calleau! Callai! | Wed Nov 12 1986 16:27 | 10 |
| re .13:
Why can't you re-boot the ship's computer? When I was at Nasa, we did
it. But then we never had satellites big enough to fit a 780 on them.
On the other hand, the people I worked with often said (to us wimps who
used M, VMS, and Twenex), "If you're using an operating system, you're
not *really* doing real-time."
Jon
|
| 351.16 | longevity | AIWEST::DRAKE | Dave (Diskcrash) Drake 619-292-1818 | Sun Nov 16 1986 21:31 | 5 |
| As an OEM we had a 780 stay up well over 150 days. You should have
seen the queue job numbers! The fan on the LSI-11 took it down.
We ran 3.x circa 1981-1982. I agree with the note that suggested
"Methusala" testing for longevity, you never know what evil lurks
behind the curtains...
|