I’ve written a sched. agent. Its in LS, and it basically checks some files on the network…if the file exists it sends an email to certain users, if it doesnt it sends a file to other users. Very simple.
When I click on Run, or test, the agent works, and it send out the e-mail. However, when it runs during its ‘schedule’ the e-mails are never going out.
How can i debug/troubleshoot this? The agent is signed properly with correct rights.
My question is, why do agents always have to run on certain amount of documents! can’t we just run ‘code’ at a scheduled time/date and not have to be forced to choose new or all all documents?!
Make sure to run “test” diagnostic! It will double check your assertion about rights and scheduled server. Look at server log for errors. If agent runs and mail is sent it is likely to be a routing problem. If agent does not run, it is likely to be something with your agent rights or usage of UI functions.
Thank-you all for your suggestions: here was the solution.
1st I was diming and setting the ui workspace (not allowed on server agents);
2nd I was doing some network operations in LS and therefore in the agent properties under security tab I had to select option 3. Allow Restricted Operations with full administration rights.
Im not sure but I think it could be something to do with the fact that scheduled agents run on the server rather than when you click run from the agents list it runs on the client. this would particularly effect your instance as you are checking files on the file system. bit of a long shot but just an idea
The agent is signed by the server correct? Did you try using an Error routine to see if the script produces an error? Or does the agent just fail to send mail?
andre, ur right ‘all documents’ is not intuitive at all.
“You need to make sure that the server is properly set up to be able to route email. Otherwise, no mail sent by agents will be delivered.”
…my server is a Mail server…it better be properly setup! its sent mail before using other scheduled agents, only difference is that the other agents were written in formula and this one is in LS.
micheal, i havent tried using an error routine to see where the problem actually lies. It could be bombing out way before the e-mails are trying to be sent.
You run Domino as a service under Winbloze. Winbloze has a “feature” whereby nothing run as a service can access mounted (mapped, whatever) network resources. Only interactive logons can use them.- I see under XP Pro, the services has a tab called “Logon”. I suspect this allows a workaround were Domino can be run as a particular account, thereby making it “interactive”, so it can in fact get to network resources. I’d try setting that, if you have it available, and see if your scheduled Agent works.
If you are not running Domino as a service, or you are and you don’t have the Login tab, and you’re unwilling to have the computer logged in to some account 24x7 to launch Domino interactively, you’re pretty much hosed. At one time, during the R4 time frame, I tried to write a C++ launcher to run Domino using a specified security context under NT. MickeySawft makes this INCREDIBLY difficult, so it was ultimately abandoned. Your mileage may vary.
Try Linux. (grin) I ran R5 on Linux as a development platform for years; I’m sure R6 is equally effective under that OS, and it doesn’t have this bizzarro network restriction. Domino does have limitations under Linux that are not evident under Winbloze, however. Some are obvious, like no COM/ActiveX support, which is a Good Thing ™. Much safer, but embedded Excel won’t work. (shrug) Some are not so obvious, such as certain elements not working, which are pretty much laid out in Help. I didn’t bump into any of those with the R5 setup.
When I click on Run, or test, the agent works, and it send out the e-mail. However, when it runs during its ‘schedule’ the e-mails are never going out.
When you run the agent on your client, it’s using your mail server to send the mail. When you run it on the server, the agent doesn’t have information about your Notes client’s mail routing setup. You need to make sure that the server is properly set up to be able to route email. Otherwise, no mail sent by agents will be delivered.
How can i debug/troubleshoot this? The agent is signed properly with correct rights.
Please see this article for tips on troubleshooting server agents.
Your assertion that the agent has “correct rights” is not convincing. How are we supposed to know that you know what the correct rights are? Please be specific.
My question is, why do agents always have to run on certain amount of documents! can’t we just run ‘code’ at a scheduled time/date and not have to be forced to choose new or all all documents?!
Just choose all documents – selecting all documents takes no time, since the database already has a collection of all the documents that the agent can just point to. There’s nothing that says that your agent has to even look at the UnprocessedDocuments collection.
Yes, it might be a little more intuitive to have a “None” selection, but there’s no practical difference.