I would need more complete information to properly debug all of these individual issues effectively which would include at least....
-Java version
-OS name w/ version
-HPS Java version
-exact input file you are using (an absolute path accessible from ifarm or jlabl3 is preferred rather than just a file name)
-whether you are running over ET or file streaming
-if ET whether you are using the actual DAQ ET or the Java-based one that comes bundled with HPS Java
-the .prop file you are using
Any other special considerations or things that are broken in the GUI.
Both file streaming and the "mock" ET ring are working fine on my Mac right now with the trunk on EVIO files with ECAL data. I checked EcalMonitoringFinal and all the plots show properly, etc.
In fact, you're not likely to do much better going "backward" so let's try to solve the issues you're having right now, and I'll patch up the code as necessary.
For now, can you put hpsecal_004466.evio.0 and hpsecal_004469.evio.0 someplace accessible where I can get them easily e.g. not behind the hall gateway? This way we will be testing on the same files.
Hopefully, going to set up 2-factor auth tomorrow for using VNC. Have an open ticket with support.
________________________________________
From: Maurik Holtrop <[log in to unmask]>
Sent: Thursday, April 16, 2015 6:01 PM
To: McCormick, Jeremy I.
Cc: hps-software
Subject: Re: Monitoring App - Issues
Hello Jeremy,
Thank you for the detailed reply. I am using the 3.2.1 release. Same jar for Linux or Mac. To fully thorough, I tried a local maven compiled jar AND separately the jar from Nexus. Results are identical.
I tried on Centos 6.5 in a virtual machine over X. My first observation is that, no indeed, Java is not independent of operating system. Several behaviors are different in Linux from Mac OS. The key was reading from a file. In Linux, when I select a file from Open… it just opens the file and I can connect. On the Mac, I first must choose “Evio” as file type, then select an evio file, then I can connect. The GUI looks quite a bit different on the Mac, with the window sizes all wrong, but that is no big deal
Both monitoring session take a while, but eventually open the file and load conditions. Both then sit forever at:
Thu Apr 16 20:38:26 EDT 2015 MonitoringApplication startSession
INFO: new session successfully initialized
So far 2800+ seconds for Mac and 450+ seconds for Linux, so I have been patient.
I can send events (pause, go) to the GUI and they are received, it is just that no events are actually processed.
I have tried EcalMonitoring.lcsim and EcalMonitoringFinal.lcsim steering files, and I have tried hpsecal_004466.evio.0 and hpsecal_004469.evio.0.
At this point I think I have tried all possible things I can think of to make this work, including being patient (hard for me). It’s now taken just over 2 hours.
Please debug this thoroughly or tell me what I should do. Tomorrow after the MCC meeting I will go to the counting house and try there, and revert back as far as I need to go to get something that at least limps along.
Best,
Maurik
> On Apr 16, 2015, at 7:15 PM, McCormick, Jeremy I. <[log in to unmask]> wrote:
>
> Hi, Maurik.
>
> This is in reply to several emails you sent to me today about issues with the monitoring app.
>
> CC’ing to software list in case other are having similar problems….
>
>> For org/hps/steering/monitoring/EcalMonitoringFinal.lcsim no plots are shown, no ECAL event display shows up. I am using the detector: HPS-EngRun2015-Nominal-v0. The results for EcalMonitoring.lcsim are similar.
>
> This isn’t something I was able to reproduce locally either with direct streaming of an EVIO file or using an ET ring. I can see the ECAL plots fine.
>
> Is it possible that LCSim is still initializing when the plots are not showing? How long are you waiting for the plots to show up? Sometimes it takes a little while to load all the conditions data before LCSim gets setup.
>
> If you look at the console, are there a bunch of messages scrolling by about conditions system initialization?
>
>>
>> There does appear a mysterious white box, with a bar above it that says “click on a plot region” but it doesn’t do anything.
>
> This is the plot statistics that shows in the lower right hand panel. If you click on a plot (when they are showing) then it will allow you to see statistics for objects in that region like Histogram1D, etc. But this is entirely separate from the plots themselves which will show in the window about it.
>
>>
>> Issues with settings file:
>>
>> SteeringType=RESOURCE ## This seems to be ignored and default to FILE.
>
> Yes, this was some kind of bug where this setting wasn’t reflected in the job settings. It should be fixed in trunk now I think, and RESOURCE is showing correctly now for me.
>
> You also wrote….
>
>> Opening a file doesn’t work for me. I get a "Exception in thread "AWT-EventQueue-0” java.lang.RuntimeException, followed by a “failed to create ET connection”.
>
> In order to use files, you go to File > Open and select an EVIO or LCIO file. Then it should be showing as the current record source in the combo box at the top of the GUI window.
>
> Then when you click on the connect button this file should be used for the session.
>
> In this case, I think somehow you have an ET system selected instead as the event source. Instead you would want to select the file source from the drop down box.
>
> There is also an event source for the current ET ring settings that will show as “ET@localhost” etc. which can be used to select the ET source.
>
>> Also I get no trigger diagnostics on the window even when using the TriggerDiagnosticsMonitoring.lcsim
>
>
> There is a config event every 10k events, but all of them will not be found in the data. So you may have to wait a little while for the trigger diagnostics to initialize. This is a bit haphazard but we found in the past that attempting to do this initialization asynchronously did not work correctly. If the trigger diagnostics do not initialize after awhile (not sure…a minute or two?) then something else is probably going wrong.
>
> —Jeremy
########################################################################
Use REPLY-ALL to reply to list
To unsubscribe from the HPS-SOFTWARE list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1
