January 2010 - launchd-dev - lists.macosforge.org

Enforce a maximum time that a launchd process can live...
by Michael Ledford 26 Jan '10

26 Jan '10

Hello, I have a process I would like to used launchd to control. Its job is very simple; look for messages that are waiting in a SQLite database and send them synchronously with NSMutableURLRequest via HTTP. The problem I face is occasionally on production machines I have found that the networking request becomes stalled for some reason which causes the process to simply sit waiting. I was hoping that launchd would be able to enforce a maximum time that a process can live using the ExitTimeout key. The documentation isn't very clear to me on this point. At first I took it as this is the maximum length of time the process would live before launchd would send it a SIGKILL. However, it was pointed out to me by someone they believe it is the time between launchd sending a SIGTERM and SIGKILL. This would only occur if the launchd process was stopped or removed and doesn't exit within the specified ExitTimeout causing a SIGKILL to be sent. In the ideal world people would say, "Your problem is a bug and it needs to be fixed." However, in the real world where frameworks are not under my control this isn't an option. In this case I want to work around the problem. I want to specify a maximum amount of time that a process can live and then have it terminated if it exceeds that time. Can I accomplish a process living for a certain amount of time and then being asked to terminate via launchd? If not, what other options do I have? Thanks, Michael

2 3

Re: [launchd-dev] SMJobSubmit( ) fails but no error is returned
by Karl Moskowski 20 Jan '10

20 Jan '10

Thanks to Damien for finding the problem. In case anyone else encounters the same thing, here's the summary. During development, I had manually unloaded the agent using "launchctl unload -w". This had the effect of setting a disabled key under the job's identifier in the launchd database at: /var/db/launchd.db/com.apple.launchd.peruser.NNN/overrides.plist (where NNN is your user ID number, e.g. 501) With that setting in place, SMJobSubmit() returned false, but the outError argument was NULL. Manually editing overrides.plist to remove the offending entry allowed SMJobSubmit() to work. I've filed rdar://7561469 in case anyone else wants to file a bug. ---- Karl Moskowski <kolpanic(a)voodooergonomics.com> Voodoo Ergonomics Inc. <http://voodooergonomics.com/>

1 0

Re: [launchd-dev] SMJobSubmit( ) fails but no error is returned
by Karl Moskowski 20 Jan '10

20 Jan '10

I've narrowed it down a bit. SMJobSubmit( ) fails without an error only on one user account on my work iMac. If I switch to another account, it works fine. The ~/Library/LaunchAgents/ folder looks OK. Maybe something in my home directory is corrupted or has the wrong permissions or ACLs. Any idea where to look? TIA again. ---- Karl Moskowski <kolpanic(a)voodooergonomics.com> Voodoo Ergonomics Inc. <http://voodooergonomics.com/>

2 1

SMJobSubmit( ) fails but no error is returned
by Karl Moskowski 15 Jan '10

15 Jan '10

I've been trying to use SMJobSubmit( ) to have launchd run my LaunchAgent. The weird thing is, it works on my MacBook Pro at home, but on my iMac it fails - the return code is false, but the outError argument is NULL. Both machines are running up-to-date Snow Leopard and using Xcode 3.2. Source for the app (a Sparkle utility to check all apps for updates) is at <http://bitbucket.org/kolpanic/coruscation/>. I'd appreciate any insight. I create a NSDictionary corresponding to the plist, then submit the job with: CFErrorRef *error = NULL; if (!SMJobSubmit(kSMDomainUserLaunchd, (CFDictionaryRef)plist, NULL, error)) NSLog(@"Error in SMJobSubmit: %@", (NSError *)error); if (error != NULL) CFRelease(error); Below is an example of the plist saved to ~/Library/LaunchAgents/. If I open Terminal and manually use launchctl to load the plist, it works. <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version="1.0"> <dict> <key>Label</key> <string>com.voodooergonomics.CoruscationAgent</string> <key>Program</key> <string>/Users/karl/Projects/build/Debug/Coruscation.app/Contents/MacOS/CoruscationAgent.app/Contents/MacOS/CoruscationAgent</string> <key>RunAtLoad</key> <false/> <key>StartCalendarInterval</key> <dict> <key>Hour</key> <integer>9</integer> <key>Minute</key> <integer>14</integer> <key>Weekday</key> <integer>5</integer> </dict> </dict> </plist> ---- Karl Moskowski <kolpanic(a)voodooergonomics.com> Voodoo Ergonomics Inc. <http://voodooergonomics.com/>

2 2

Launchd sanitizing python environment
by Heath Carlisle 01 Jan '10

01 Jan '10

Hello, I'm a relative noob to launchd. Thanks in advance for any help. 10.5.8 , plist run as a User LaunchAgent I am running gimp batch jobs via a wrapping python script. When I run the wrapper from the command line (bash), all is well. When I schedule it via launchd, the environment appears to become sanitized and gimp cannot be found.(I'm comparing environments by issuing the "env" command from the wrapper script.) I can set, for instance, the $PATH via the plist <EnvironmentVariables> key,(which will then find gimp) but then other dependant environment variables cannot be found, such as the display. Since the display (and others) are /tmp/ items, they cannot be "hard-wired" into the plist. As gimp requires a host of helpers(X11, GTK+, GEGL, PANGO etc), debuging these one env variable at a time seems monumental. Is there a way to just run the job as if it were from the user's shell? Perhaps I've overlooked some basic concept? thanks, -hc -- View this message in context: http://old.nabble.com/Launchd-sanitizing-python-environment-tp26799801p2679… Sent from the Launchd mailing list archive at Nabble.com.

2 4