Ready to Start... but nothing happens
25 Feb 2016 2:36:33 UTC
Topic 198465
(moderation:
2 PC's of same make and model - install Boinc Manager for Linux (Mint 17.2) and join a few projects (same on both PC's) One PC works fine and the other says Ready to Start or Waiting to Run but just kinda sits there. Rebooted several times - R&R Boinc - R&R VirtualBox... and it just stares at me.
Ready to Start... but nothing happens
)
Hi Ken,
Your list of computers shows 7 machines that have had contact with the servers in the last week or two. It would be helpful if you could point out exactly what machine you are referring to.
Exactly what science runs have you selected in your preferences? If you are trying to get work for FGRP4, that could be the problem since that run has essentially finished and is replaced by FGRPB1. I tried to look at machines where you had two of the same type. There appear to be two separate pairs.
If you are talking about the hosts with i5 520 processors, each has a single completed task and no additional tasks. Is work fetch allowed on both of those?
If you are talking about the T9400 core 2 duos, one has the new FGRPB1 tasks whilst the other has no tasks whatsoever. In its last scheduler contact on 16 Feb, it was not asking for any work which suggests you have turned off work fetch on that machine. Maybe it's the same machine as the other but with a different host ID. Check out which particular ID you are referring to out of the 7 in the link above.
If you restart a problem machine and copy and paste the startup messages in BOINC Manager into a message here, someone might be able to tell you what is going on.
Cheers,
Gary.
Wow, thanx for the quick turn
)
Wow, thanx for the quick turn around!
Hi Ken,
Your list of computers shows 7 machines that have had contact with the servers in the last week or two. It would be helpful if you could point out exactly what machine you are referring to.
There are 4 HP laptops I pulled from the recycle pile at work and put together with duct tape and scrape parts. They are named Ken-Boinc-1 through 4. I'm Hoping to have a 5 and 6 when we roll out some new replacements. I'm thinking the culprit is number 1 or 2 but I will verify when I get back to work.
Exactly what science runs have you selected in your preferences? If you are trying to get work for FGRP4, that could be the problem since that run has essentially finished and is replaced by FGRPB1. I tried to look at machines where you had two of the same type. There appear to be two separate pairs.
They are all processing Atlas, Einstein, LHC, Milkyway and SETI. I'm not familiar with FGRP4/B1. I don't recall setting a pref or even if there is a pref to set this.
If you are talking about the hosts with i5 520 processors, each has a single completed task and no additional tasks. Is work fetch allowed on both of those?
No, this is too new a PC. This may be my wifes PC.
If you are talking about the T9400 core 2 duos, one has the new FGRPB1 tasks whilst the other has no tasks whatsoever. In its last scheduler contact on 16 Feb, it was not asking for any work which suggests you have turned off work fetch on that machine. Maybe it's the same machine as the other but with a different host ID. Check out which particular ID you are referring to out of the 7 in the link above.
These sound like older PC's so this may be the place. Again I will check when I get back to work. I am at home now and trying to open the Read Config and Read Local Prefs files. And a simple click on the menu option isn't doing the job. Thoughts? Where/How can I set the prefs to be the same across all my PC's? Such as... always fetch new work when a network connection allows. Plz use small words as XML files and basic programming is all Witchcraft and Sorcery to me! :)
If you restart a problem machine and copy and paste the startup messages in BOINC Manager into a message here, someone might be able to tell you what is going on.
Where does one get the start-up msgs? Are we talking about the BOINC Event Log files? Sorry about the techno ineptitude but I am just trying to help the cause... and have a little fun. BTW, Thanx for all your efforts. Personally, I think the BOINC program as a whole, regardless of which science project one tends to associate with, is a Noble effort.
RE: They are named
)
Only you can see the names of your computers, that's why Garry asked for the hostIDs. You can see a list of all your hosts by going to "Your Account -> Computers on this account" click on the "View" link.
By science runs Garry is referring to the different searches done by Einstein@home, not other Boinc projects. You can select what science runs you like to participate in by going to "Your account -> Einstein@home preferences" and then choosing "Run only the selected applications" when editing the prefs.
Those are options make Boinc read the files again to pick up changes made to them, you can't open the files to read them that way.
Use the two preferences pages here on the website. Go to "Your account -> Computing preferences" to set up how Boinc will use your computers, ie how much work to download. The Computing preferences are global and will propagate to the other projects your running. Set them at one project only.
The project specific preferences can be found at "Your account -> Einstein@home preferences" and these are to setup more specifically how Einstein will "use" your computers.
Yes, Garry is referring to Boinc's Event log, if you restart Boinc and then open the event log ("Tools -> Event log" in Boinc Manager), select the first 20-30 lines and then paste them here and we will have a better understanding of whats going on.
Hi Ken, Holmis has already
)
Hi Ken,
Holmis has already covered the most important things so I'll just concentrate on a couple of other points for the moment.
Firstly, apologies for looking at your join date of July 2005 and assuming you would have a certain level of familiarity with current workings. It's hard for people like me to really remember how difficult things are when you first try to understand it all and how overwhelming it can be when people make bad assumptions :-).
I'm really pleased you're back here at a very exciting time - the first direct observation of gravity waves. This is an event that's been 100 years in the making. What you are doing - rescuing otherwise defunct machines and putting them back to useful work is laudable. Whether you like it or not, by your very own admissioms, you are already in the exclusive geek/nerd club and it is incumbent upon you to tweak configs and edit .xmls as you unlock the full power of BOINC :-). So, you just need to learn the lingo ;-).
Your first task is to understand the different science runs (sub-projects if you like) that Einstein hosts so you can select what interests you the most. You will find a summary of them here, amongst quite a few other things, hopefully useful. In understanding about different projects and their science runs, it's useful to know there are three basic components at the user end of the full BOINC system, (1) the BOINC client itself which does no crunching, but handles all communication with projects and the running of (2) the project applications which do all the work. Above this, sits (3) the BOINC Manager which is just a GUI interface to the BOINC client. BOINC itself is very happy to never have to talk to the Manager (as are most conscientious workers ;-) ) since it has built-in basic workable settings and it can always get instructions from the website whenever it talks to the project's servers. The real purpose of BOINC Manager is just to allow you (the CEO if you like) to meddle with things instantly and potentially stuff things up whenever you so desire :-).
Your second task is to study the website preferences mechanism for global changes to preferences. You access these through your 'account' page on the project website. 'Global' simply means that the changes will apply to all hosts for all projects that are part of the BOINC ecosystem and that it might take a while for these to propagate. You need to fully appreciate how global preferences work and the differences there will be if you override global settings with local settings available through BOINC Manager on individual machines. Once again there is some information in the above link.
While you are exploring the various types of global settings through your account page, also please check the link for viewing all computers. This page shows all the hosts on your account. In the Computer ID column you can see the assigned host ID and the links for 'details' and 'tasks'. Make sure you explore these.
Your third task is to become familiar with the BOINC Manager - Advanced view interface. A long time ago, the BOINC Devs decided that this was too 'frightening' for the average user and made the 'simple' view the default one. If you are serious about wanting to understand what's happening and why, you need the advanced view. Click through all the tabs and drag the column widths to be want you want, so you can see the information you need. If you have a wide enough screen, it can be useful to put the main window and the event log window 'side by side' so when you make some change you can read conveniently what is reported in the event log. If you have the time and want to see the full scope of what's available, please consult the Advanced view page in the BOINC User Manual. As your experience grows, stuff in the User Manual becomes more and more useful.
There are two of these i5-520 hosts showing in the list of your hosts. One is producing results but the other has no work in progress at all. I suspect the one physical bit of hardware may have acquired two different host IDs. If so, it's not really a problem as they can be merged later if necessary.
There is also a (more modern I suspect) i5-2540, which seems to be sailing along just fine. Maybe this one is your wife's machine.
The T9400 that had work and was crunching is now returning compute errors. You can see the error messages by going to the list of tasks for that host on the website and clicking on the task ID of one of the tasks with the error. In brief, the error message is, "Too many exit(0)s". BOINC regularly monitors the science app. If it exits prematurely, but without an obvious error, (exit (0) ), BOINC will restart it. It it exits with a real error (non-zero exit status), BOINC will clean up immediately. There is a limit to BOINC's patience so if a science app gives a premature zero exit too many times (100 times I think) BOINC will 'pull the plug' at that point.
Perhaps the machine is running too hot. Is it in a closed office with no aircon? This could be the cause. You should be aware of one big difficulty with using laptops is keeping them from overheating. This is a particular problem for older machines where the relatively narrow airways may have been partially clogged with fluff and dust or where the cooling fan has grown weaker over time. Laptops were never really designed for 100% duty cycle crunching.
As you read the suggested stuff and explore the methods of controlling things, please ask about anything you don't understand. There are many people here willing to have a go at answering such questions. I look forward to hearing more when you get to work and can see the status of those machines.
Cheers,
Gary.