Messages

Yep, a combined "pause and resume" action is what I was thinking, but we have to account for the occasional task that is thoroughly messed up to where that "reset" doesn't fix it.  In those instances, after we reset it and it still doesn't finish, then we need to either permanently suspend or abort it - which I was assuming would be a second rule.

Fred -
I can't begin to express what a benefit these rules have been.  This truly is a game changer in terms of automating boinc.

OK, after playing with it for several weeks, there are a few things I'd like to propose. 

The first (and most important to me) is a new "pause & resume" action.

Especially for DNETC, tasks will get "stuck" but they are reset when you suspend and then resuming them so that usually they will finish successfully.  Right now, I'm just suspending when they get to a time mark and then manually restarting them later.  I'd love to have a "pause & resume" action within the rules so that I could automate doing that.  The key here is that I need to be able to have 2 rules with different thresholds so when a task gets to 8:00 it will reset (and its fine if it does it every minute) but once it gets to 12:00, I need the second rule to engage that suspends (or eventually aborts) it entirely.  Are rules processed in the order they are entered?  So that I would need to enter the 12:00 rule first and then the 8:00 rule, at 9:00 will it fail the 12:00 rule and then process the 8:00 rule?  Or would it process both and therefore the second rule would resume the pause that was intended to be permanent?  If that were the case, I'd really need an abort action to be added now for this to work.

I'd still like to see a rule that fired when the error count exceeded a trigger point on a project within a given time period.  So, if I had 10 errors on primegrid within a 5 minute period, as an example, then I could suspend that project.

And the other rule that I'd still like to see is to have a rule triggered when a machine loses its connection and fails to reconnect.  I'd only use this one with the external scripting capability to reboot the box that wasn't connecting (and was presumably frozen).  It's doesn't happen a lot and therefore is a lower priority, but this rule would really allow a bulletproof operational environment.

Thanks again for all you are doing here.  I keep BT up 24x7.  Good stuff.

The "Progress % / min"  rule event is exactly what I needed, but though it isn't really using a minute as its sampling and therefore it is firing prematurely.  DNETC only updates the % complete every 15 seconds or so, but then it will jump a couple of %.  Right now, BT is noticing that it goes 3-5 seconds with no progress and then is extrapolating that to be 0% per minute, which is causing the rule to fire when it shouldn't.

Is this tied to the refresh rate?  Any way we can make this rule sample over a longer time period?  Would prefer that it be a full minute between comparisons.

In 0.63 changed to a time base of about 1 minute instead of the 0.5 minute as it is now.
The time window will move along at a pace of about 15 seconds.
I will test this on DNETC@Home

And the test..... Another project that is not using the progress at all. Stays at the same progress forever.
Then jumps over to another value and stays there forever.
So the rule won't work because there is no progress at all, for a long long time.
BoincTasks shows progress, because it uses a fall-back mode, when progress is close to 0. It uses the Time left to get a progress anyway.

The only way for this rule to work is to set a very long time like 45:00 (45 minutes)

Drat!  That wouldn't work.  OK, for now I guess I'll just have to monitor total time elapsed.

But the project does report time left?  I wonder if that could tell me something.  I haven't been watching that closely.  Guess I need to start paying closer attention and see if we could do something like time left reduction / minute.

Thanks.

Oh.  OK.  I understand now.  So this only affects the local computer (which is why the "localhost" config is so important for this feature).

The "Progress % / min"  rule event is exactly what I needed, but though it isn't really using a minute as its sampling and therefore it is firing prematurely.  DNETC only updates the % complete every 15 seconds or so, but then it will jump a couple of %.  Right now, BT is noticing that it goes 3-5 seconds with no progress and then is extrapolating that to be 0% per minute, which is causing the rule to fire when it shouldn't.

Is this tied to the refresh rate?  Any way we can make this rule sample over a longer time period?  Would prefer that it be a full minute between comparisons.
It is now sampled over 28 seconds, you can set a time. This will cause the rule to trigger only if the rule stays valid for a longer period of time. E.g. 5 minutes.

Sorry to ask an even dumber question, but where are the snooze and snooze GPU buttons?  I don't think I've ever seen them.

Did you not say that I had to enable history for the new rules to work?

Being able to keep just a few hours is likely the best option for me. 

And as you can see, I have two messages there.  The one on top is actually a stack of perhaps a dozen, so it definitely gives the warning more than once.  Perhaps once an hour???

As discussed previously, I am still having a problem with the history file becoming too large.  It is currently set to 1 day.  Screen cap with exact error message attached.  Let me know if there is any additional diagnostics I can provide.

3) If you need a rule to apply to a certain computer, best use the right mouse button in Tasks and select add Rule.
This way you don't type the wrong name.

This doesn't work for me.  Are you certain it is tied to the column regardless of positioning and still functions when I move the "computer" column?  Because when I right click in the tasks window, nothing happens.

John,

I've tested this and I can't reproduce your problem  I've moved the Computer column around to several locations including column 1 with same result: Rt-click/Add_rule calls up a rule editor on the highlighted task and auto-inserts computer name, project & application.  (BTW the test rule also works.) This is on a WinVista_x64 SP2 host.  I believe you are running on Win7_x64?

I am running Win7_x64, but here's the really weird thing... I can't reproduce it now either.  It works just fine now.  ?  Oh well, if it happens again I'll try to figure out why, but for now just ignore this one.  Sorry.

Oh, and one other thing... the rules are working great now.  I've already caught a real error and prevented it from keeping my system hung up doing nothing.  Some might wish that you could manually start a process running again with it getting throw back to suspend (ie, there is no override without disable the rule entirely).  But there are a lot of other things I would prefer before worrying about that.  Overall, very happy to have this working.

3) If you need a rule to apply to a certain computer, best use the right mouse button in Tasks and select add Rule.
This way you don't type the wrong name.

This doesn't work for me.  Are you certain it is tied to the column regardless of positioning and still functions when I move the "computer" column?  Because when I right click in the tasks window, nothing happens.

5) The history option MUST be switched on.

Doh.  There's my problem.  I tried enabling history once before, but I ended up getting some kind of out of memory error (on a 16GB machine) even with it turned down to just one day of history.  I'll try it again and test the rules as well as try to determine what exactly the issue was with the history.  Thanks.

Shouldn't the "messages" roll off after some period of time?  Once you get a dozen machines connected, I have to believe all of that stuff adds up and it clearly isn't following the config for history.  I'd have a separate config on the history tab asking how long to keep messages.  Default in my mind would be two days.

Also, even one day of history is too much for me and since it is a drop down, I can't include fractional values.  Should that start with a couple of entries such as 1 hour, 4 hours, 12 hours, before counting up numbers of days?

Since we don't yet have "application stalled", I just created a rule for computer "Cruncher-1" on project "DNETC@HOME" (no application) that if Elapsed Time > Value(12:00) (No Time) then color purple and suspend task.  I've had a couple tasks go over, but nothing has happened.

1.  Are the rule actions not coded yet?
2.  Is application or time required?
3.  Did I somehow misspell the computer or project name (I am sure they both match what is shown on the tasks screen)
4.  Is there a bug?

Not sure what is wrong, but I'm not seeing what I was expecting to see.

Love the new rules area.

I have a problem with the minute field (Middle field) of the timer field.  What ever I place in the second field ends up in the third field.  Therefore I am  unable to specify minutes.

A problem that I can't reproduce for some reason.
Try something like 0d,00:30:00 or 30:00

I also am not seeing this problem, either in the Time or the Value field of the Rule Editor dialog

Odd.  Like you (but unlike the others) I am not having this problem.  For whatever it is worth, I am running Win 7 x64 bit.

I will add  the possibility of 1 internal and 1 external event. So starting a program and suspending a task.

Like it.  Thanks.