Re: Changing data structure forced command termination.

Jacob Salomon wrote: > > Huh? > > Hi Family. > > One of my users just upgraded their engine to ODS-7.24.UC3 from > 7.13.UC5. Of course, no upgrade is without some grief. Here is a > manifestation of the grief we get now: > > I ran onstat -u and got some lines of output - all the daemons and the > first bunch of users in the sessions list. Suddenly the output ends, > way before it should, with the message: > Changing data structure forced command termination. > > I tried onstat -g ses and got exactly one line of output before getting > the same message. > > THIS IS NUTS!! If the data structure has changed, the onstat code was > surely changed to match it. If the code was not sufficiently changed, it > would not have gotten a single line out the door; it would probably have > dumped core. > > I don't know what other commands will be terminated with this error > message. So far, onstat -g ppf, ath, act (besides the ses) have all > ended with this error message. I don't need to try every possible > option before realizing there is a more global problem here. > > The app seems to running fine without recompilation (OK, some termcap > problems seem to be cropping up.) > > Anyone got a clue? Jacob, There is some bug (try #68290 with tech. support) wherein if either the host or tty name as stored in syssessions (or sysses-whatever) is greater than 8 characters, then onstat -u/-g ses will freak out and dump this error on you. It was supposed to have been fixed as of 7.23.UC5. I saw this still occurring (in my case, the tty and the host name as reported by Win95 were both > 8 chars - this happens often when using DNS as often the full domain name of the machine is sent to the server as the host name) in 7.23.UC4. Good luck, Wes Gamble