I didn't know if this got lost in the threads, as it was started a long time ago, so I thought I'd push it up to current by starting a new thread.

: I don't think this will shed any light, as you already know it will stop at
: some point.
: I would like to suggest setting all jobs to run detached. This should free
: the queue. Also set different jobs to run in different queues to prevent
: them from blocking each other.

: PS. The reason you don't see any errors in the log is that the logging is
: probably tuned off for these jobs, this is a default for frequently
: running jobs. If this is the case turn it on while you are troubleshooting
: the issue..

I have upgraded my hardware and 24x7 since the last reply to this thread, but have not had any luck. Nothing in the 24x7 logs (I made sure logging was turned on) or windows event logs. The "Keep ALive" process I made checks for a file that is never deleted and writes the date/time to that file, so I can pin point when it happens, but no other info. I have all jobs running as detatched and in their own queues. It could go for days before acting up, but it never happens more than once in 24 hours. In other words, if it acts up and I restart 24x7, it will not happen again for atleast a day, usually not for a week or more.

It is still unclear what is happening. Does 24x7 run other jobs or the entire thing hangs?

Maybe we should start from the beginning:
You say "Service stops looking for semaphore file." Does that mean other jobs continue running but not this one or just nothing runs? In other words it just ignores the file or cannot find it or does not look for the file because it is busy doing some else or hung?
Have you tried scheduling a test job (to see if it will continue running) as suggested in the old thread? If yes, what have you found out?

Can you clarify this situation so we know where to look for the root cause of the problem?

: I didn't know if this got lost in the threads, as it was started a long time
: ago, so I thought I'd push it up to current by starting a new thread.

: I have upgraded my hardware and 24x7 since the last reply to this thread, but
: have not had any luck. Nothing in the 24x7 logs (I made sure logging was
: turned on) or windows event logs. The "Keep ALive" process I
: made checks for a file that is never deleted and writes the date/time to
: that file, so I can pin point when it happens, but no other info. I have
: all jobs running as detatched and in their own queues. It could go for
: days before acting up, but it never happens more than once in 24 hours. In
: other words, if it acts up and I restart 24x7, it will not happen again
: for atleast a day, usually not for a week or more.

: It is still unclear what is happening. Does 24x7 run other jobs or the entire
: thing hangs?

: Maybe we should start from the beginning: You say "Service stops looking
: for semaphore file." Does that mean other jobs continue running but
: not this one or just nothing runs? In other words it just ignores the file
: or cannot find it or does not look for the file because it is busy doing
: some else or hung?
: Have you tried scheduling a test job (to see if it will continue running) as
: suggested in the old thread? If yes, what have you found out?

: Can you clarify this situation so we know where to look for the root cause of
: the problem?

24x7 is used only for publishing in the fashion outlined:
A semaphore file is created, 24x7 checks for the file once a minute, when it finds one, it starts a job based on the file it finds (they are all syncFTPdir() jobs).
Publishing different websites requires differently named semaphore files. 24x7 just stops looking for the file. When I publish, the file is created, the service is running, but it never "sees" it, so the job is never run. None of the jobs will run when it starts doing this, but then again, they all run on semaphore conditions. It is not hung, because I can run it from the user interface and it is not busy doing something else. I have also place all the jobs in their own queue and have them run detached. As stated before, there is nothing in the 24x7 logs or in the Win2k event logs (app, system or security).

I did create a job that looks for a semephore file every 15 minutes. When it finds it, it writes a timestamp to it, but does not delete it. This creates a file with time stamps every 15 minutes. When it stops working, it is reflected in the "keep alive" log. I will change this job so that it does not look for a file and just writes to the log and see if the 24x7 stops running jobs altogether or if it just ignores semaphores. I'll post results.

Thanks for the update.

Just in case, a single line JAL script job with the following script can be used as "heart bit" or "keep alive" job

FileSave "c:\\24x7bit.txt", "@T"yyyy-mm-dd h:mm:ss""

: 24x7 is used only for publishing in the fashion outlined: A semaphore file is
: created, 24x7 checks for the file once a minute, when it finds one, it
: starts a job based on the file it finds (they are all syncFTPdir() jobs).
: Publishing different websites requires differently named semaphore files.
: 24x7 just stops looking for the file. When I publish, the file is created,
: the service is running, but it never "sees" it, so the job is
: never run. None of the jobs will run when it starts doing this, but then
: again, they all run on semaphore conditions. It is not hung, because I can
: run it from the user interface and it is not busy doing something else. I
: have also place all the jobs in their own queue and have them run
: detached. As stated before, there is nothing in the 24x7 logs or in the
: Win2k event logs (app, system or security).

: I did create a job that looks for a semephore file every 15 minutes. When it
: finds it, it writes a timestamp to it, but does not delete it. This
: creates a file with time stamps every 15 minutes. When it stops working,
: it is reflected in the "keep alive" log. I will change this job
: so that it does not look for a file and just writes to the log and see if
: the 24x7 stops running jobs altogether or if it just ignores semaphores.
: I'll post results.

: Thanks for the update.

: Just in case, a single line JAL script job with the following script can be
: used as "heart bit" or "keep alive" job

: FileSave "c:\\24x7bit.txt", "@T"yyyy-mm-dd
: h:mm:ss""

Ok, it happened again and I think I see a relationship between my 2 issues.

It stops looking for the semaphore file because the FTPsync fails. I didn't notice the relationship before, because it seems whenever ANY site has an FTP issue, ALL the semaphore files are ignored. I will keep an eye on this, but after looking at the time stamps on the ignored pub files and looking thru all the jobs' log files, it seems to be a good lead.

So far, one site had an FTP issue and was disabled. While this explains why that site stopped looking for semaphore files, I think it may be effecting other similar jobs in some way because since then none of the other jobs will look for semaphore files.

Although I will continue to see if the 1 job being disabled due to error stops other jobs as well, I will assume that the 2 issues are more or less related and will focus on the FTPSync issue.