Plugin: Sage Job Queue (SJQ)

[nyplayer]

Slugger,

The settings still does not work in IE and now does not work in google Chrome. If I select setiings in google Chrome it sends me to the Active list. I am only able to change settings now if I use firefox.

[Slugger]

Server 135 is bad. I don't think it's causing the UI problems (which I still cannot reproduce), but it will definitely cause a busy loop so if you're running 135 then stop it and revert back to an older one.

[Slugger]

Please try server build 141. This build fixes the busy loop bug introduced in 135. Also, note that the new task timeout options were incorrectly documented. The names of the options are :MAXRUN and :MAXRUNRATIO. The wiki doc has been updated. Use of an unrecognized task option is what caused the infinite loop in 135 and it has been fixed in 141.

Please let me know if the UI issues still exist in 141. The only one I can reproduce is the Settings panel not loading in Chrome, but instead taking you to the active queue listing. I cannot reproduce any other browser bugs (FF2, FF3, IE6, IE7). The Chrome bug seems to be an issue with its history management. You can load the settings panel in Chrome by manually appending the srvConf history marker to the end of the URL (i.e. /sjq/SJQClient.html#srvConf). I'll spend a little more time investigating this later in the week, but my quick 10 mins googling the issue suggests it's a Chrome bug that I may not be able to address.

Let me know what's still causing problems in 141. First thing I'd try is to clear your browser's cache. Old copies of the compiled js files might be conflicting.

[nyplayer]

Slugger,

Freespace question... Say I have a 500gig drive for Sagetv and I tell sage to leave 10gig free... do you estimate the freespace on the full 500 or the 490 ?

[Slugger]

Originally Posted by nyplayer

Slugger,

Freespace question... Say I have a 500gig drive for Sagetv and I tell sage to leave 10gig free... do you estimate the freespace on the full 500 or the 490 ?

Full 500. I use Java's File API for all those calculations.

[Slugger]

Snapshot 142 adds the log management features and is the last major feature for the 2.0.0 release. I'm now calling the code feature complete unless someone has something else in mind?

I'm asking anyone who's been testing 2.0.0.x builds to please upgrade your client to 136 and your server to 142 and give everything a last run through. Please report any issues/bugs/etc. Add bug reports here, but even better is to add them to the project issue tracker. I prefer the latter, but will take them anywhere I can get them.

Unless I get reports from you, I plan on packaging up 2.0.0 and releasing it "officially" sometime early next week.

[jwtucker]

Any chance of getting the client to run as a service before calling 2.0 final? I'm running the client on a WHS box and don't want to leave it logged in all the time if I can avoid it.

Thanks for a great plugin!

James

[Slugger]

Originally Posted by jwtucker

Any chance of getting the client to run as a service before calling 2.0 final? I'm running the client on a WHS box and don't want to leave it logged in all the time if I can avoid it.

Thanks for a great plugin!

James

I'm going to give it a try. I've never tried to write a Windows service so I've got some googling to do, but seeing as how I'd really like my client to start up at system boot automatically I'm going to give it a shot.

Probably not until next week now. Just scored some really cheap tickets to the Dolphins/Bills game in Toronto (go Dolphins!) so my weekend's just filled up.

[Slugger]

Client snapshot 151 includes the necessary wrappers to install the task client as a Windows service. Note that starting with snapshot 151, the task client no longer accepts command line arguments for connection settings. Instead, you now need to setup a properties file with the connection settings for your task client. This was done to support automatic startup of the task client as a Windows service. Please read the updated task client installation wiki doc for further details. There are many changes to the task client setup as compared to older versions so please take the time to read over the docs before upgrading.

A new snapshot of the server is also available with some minor functional changes (namely, the entire completed task table can now be deleted - after the user confirms - instead of disabling the feature on the completed task table).

There is now just one more minor issue ticket remaining for v2.0.0, so basically these latest snapshots are feature complete. Again, I'd ask everyone who's been testing these builds to upgrade to the latest versions of the client and server and give them a final test. I'm also curious to hear feedback about the Windows service feature. Is it working? Was it easy to setup? Is the installation wiki doc clear?

[nyplayer]

Slugger,

Any reason why it looks for the properties file on the c drive. I use a central location for all my Sage Scripts and like to keep any applications away from the C drive on my clients and server.

Maybe this is needed for the service?

BTW I use the batch file because I like to delay the Start of SJQ after the Sage Service has fully started.


SJQ Task Client v2.0.0.151
Using 'C:/sjqc' as location for configuration file...
Connection refused: connect

[Slugger]

Originally Posted by nyplayer

Slugger,

Any reason why it looks for the properties file on the c drive. I use a central location for all my Sage Scripts and like to keep any applications away from the C drive on my clients and server.

Maybe this is needed for the service?

BTW I use the batch file because I like to delay the Start of SJQ after the Sage Service has fully started.


SJQ Task Client v2.0.0.151
Using 'C:/sjqc' as location for configuration file...
Connection refused: connect

There's no particular reason, I just had to pick some where to look. I originally tried using %APPDATA% and that was good until I installed it as a service and then that env var wasn't defined when running as a service. Then I tried using the 'home dir' as defined by Java, but that turned out to be just C:\ when running as a service. So I hardcoded C:\sjqc as the destination for the conf file. Choosing C: as that's the default drive on Windows and every Windows installation will have one. Then I just chose an arbitrary directory. I'm certainly open to alternative solutions here.

[nyplayer]

Slugger,

No problem I just moved the Java stuff to my central directory and kept the properties file in the C:\sjqc. I just have to remember to re-create the properties file on each client.

The reason I use a central location because I only have to update once... and each client gets the updates at once.

[Slugger]

SJQ v2.0.0 is now available for download! With this release, the 1.x stream of SJQ is no longer supported. All users are encouraged to upgrade to the v2.x stream.

A special thanks to all those forum users who have assisted with beta testing over the last couple months!

[Slugger]

v2.0.1 snapshots are now being built. No major bugs, but just things I've noticed since the release. See the delta.changes file for what's included in the builds. Very unlikely an official 2.0.1 release will be out before 2009, but I will fix bugs as I find them and you can get those fixes from the automated snapshot builds.

So far, nothing significant in these builds except for the file cleaner fix that might be of interest to those who use comskip. Basically the fix is to just add an option to completely ignore certain files. I find it useful for ignoring the *.logo.txt files generated by comskip during its runs that would otherwise possibly be deleted by the file cleaner.

All other fixes to date are just UI fixes of little or no significance (but were bugging me).

[vexhold]

Is there a place to read up on how to write rules?

I want to use this to control 4-6 instances of comskip to work on my .ts files.

Dont care when it runs or what it runs on.

Don't know where to start. Is there a beginners manual available? lol

[Slugger]

Originally Posted by vexhold

Is there a place to read up on how to write rules?

I want to use this to control 4-6 instances of comskip to work on my .ts files.

Dont care when it runs or what it runs on.

Don't know where to start. Is there a beginners manual available? lol

The project wiki area has all the docs you should need to get up and running.

I'll assume you've already installed everything and are able to access the GUI. If not, then check out the installation docs in the wiki.

For writing your rulesets, these docs should help:

• Ruleset Syntax
• Client Config Syntax

I realize you may not care if/when/why/where anything runs, but I'd suggest taking the time to setup some rules that choose exactly what is run - especially when talking about h264 (ts) files. For example, you know HBOHD is commercial free so there's no need to run comskip and waste hours processing a recording that is known not to have any commercials. Such a rule might look like this:

if [IsTV == true && IsActivelyRecording == false && ChannelName != "HBOHD"] { # Ignore HBOHD recordings
   :PRIORITY 5 # optional
   HDCOMMSKIP
}

# Use ChannelName !^ "HBO" to ignore any HBO channels (since none of them would have commercials); the !^ operator means "does not start with".  All of the operators are documented in the wiki.

Taking the time to setup a comprehensive set of rules will make the most efficient use of the hardware you're using to run your tasks. Of course, if you really don't care what it runs on or when, etc. then this rule should do it:

if [IsTV == true && IsActivelyRecording == false] { # If you want to scan while recording then remove the second test
   :PRIORITY 5 # optional
   HDCOMMSKIP

This rule will simply add the HDCOMMSKIP task to the queue for every recording after it finishes. In each client you want to run this task, simply define what HDCOMMSKIP should do:

HDCOMMSKIP {
   :CPU "NORMAL" # optional; one of NORMAL, HIGH, LOW
   :MAX 1 # Run no more than one instance of this task at a time on this client
   "C:/path/to/comskip/comskip.exe -n \"%c%\"" # The %c% is expanded to the full path of the file to be processed
}

Good luck and feel free to post questions here if you have any.

[vexhold]

The windows service will not start. I believe it has to do with me not having a C:

My main OS drive is J:

Am I correct in thinking this?

[Slugger]

Originally Posted by vexhold

The windows service will not start. I believe it has to do with me not having a C:

My main OS drive is J:

Am I correct in thinking this?

No C: would do it. I thought every Windows system had a C:? I'll whip up a quick build for you that will allow you to specify the dir name where it can find the properties file. Give me 20 mins...

[Slugger]

Snapshot 161 accepts a single parameter pointing the client to the location of the properties file. Download and overwrite your install with the new snapshot then you must add the parameter to the wrapper.conf file (found in the conf dir of the installation).

1) Stop the service.

2) Find this line in the wrapper.conf file:

# Application parameters.  Add parameters as needed starting from 1
wrapper.app.parameter.1=com.google.code.sagetvaddons.sjqc.TaskClient

And add this line below it:

wrapper.app.parameter.2=C:/path/to/props_file

Of course, substitute the value with the location where you want the client to grab the properties file from. Be sure to only specify the directory here and then create the sjqc.properties file in that directory before starting the service up again.

3) Restart service.

I've loaded this client on my system and all is running happily, but I didn't test much so please let me know if this works for you.

[vexhold]

Well, it's running now at least.

Thanks for the quick rewrite just for me... You rock!!

I will let you know if anything blows up.