From: Jamie C. <jca...@we...> - 2003-09-24 23:30:54
|
Martin Mewes wrote: > Howdy, > > Am Mittwoch, 24. September 2003 15:08 schrieb Hugo Cornelis: > > >>>If there is anything I could try again for a more >>>debugging output, >>>please let me know. >> >>I don't know anything about the webalizer module, so >>perhaps my answer is a bit naive : what I always do if >>a process seems to hang, is to use strace to attach to >>the process. >> >>e.g. if index.cgi (PID==12) seems to hang, strace -p >>12. Perhaps using the '-s' option can be usefull to >>read pending error messages in a process buffer. Of >>course this works only if you own the processes ie. if >>you are root in this case. > > > Thanks for this ... > > Well, here are the results from the "strace"-jury, to whom it may > concern: > > I initiated Webalizer-Stats with "Run on selected schedule .." > > and "strace -p 28911" as root gives ... > > write(1, "Do you want to retry the same ed"..., 36) = 36 > read(0, "", 4096) = 0 > write(2, "Enter Y or N\n", 13) = 13 > write(1, "Do you want to retry the same ed"..., 36) = 36 > read(0, "", 4096) = 0 > write(2, "Enter Y or N\n", 13) = 13 > write(1, "Do you want to retry the same ed"..., 36) = 36 > read(0, "", 4096) = 0 That does explain quite a lot actually.. When you select the new 'Run on selected schedule' option, Webmin uses a special schedule format in the crontab file like @hourly or @monthly, which is only supported by Vixie-cron. However, your system doesn't seem to be using Vixie-cron, and so is choking on that schedule format .. And clearly the module needs to handle this kind of failure better, instead of letting cron go into an infinite loop! Which Linux distribution are you running there? I had assumed that all Linux variants came with Vixie-cron, except Trustix. - Jamie |