HTTP file missing

Get physical with some of these projects? Asteroids, Cosmology, Einstein, LHC, MilkyWay, Test4Theory,
Post Reply
UBT - Chris Suddick
Active UBT Contributor 15+ yrs
Posts: 698
Joined: Tue Mar 14, 2006 12:00 am
Location: Warrington, Cheshire

HTTP file missing

Post by UBT - Chris Suddick »

Hi,

I wonder if anyone can help me?

I've been running Milkyway for some time now without problem but it seems to have gone into 'silly bugger' mode at the moment. All the tasks are completed but every time I tell it to update and get some more - it doesn't. At first I thought it was an error at the Milkyway site so I waited for it to come right but then I noticed that John's stats are showing everyone else accruing credit so I guess the problem is mine.

The Event Log shows...

06/01/2012 18:05:26 | Milkyway@home | Scheduler request failed: HTTP file not found

Any idea which file it is that it can't find?

I realise that I could reset the project but then I would lose all the credits for the tasks currently completed on my machine. It wouldn't be the end of the world but I'd like to explore other options first.

Chris.
Temujin
Posts: 2259
Joined: Mon Mar 13, 2006 12:00 am

Post by Temujin »

I've no idea really Chris Image
You could try re-starting boinc (or a reboot), maybe your DNS isn't playing fair :?

OR
post the above in the milkyway forum, if anyone can help they'll be on there


edit
I've just read the News section on the Milkyway forum and they've just moved everything onto a new server, so it could be related to that, maybe even DNS
LupusUK
Posts: 29
Joined: Tue Feb 09, 2010 12:00 am
Location: Durocobrivis

Post by LupusUK »

I tried it and I get the same error so could be a Milkyway problem
LupusUK
Posts: 29
Joined: Tue Feb 09, 2010 12:00 am
Location: Durocobrivis

Post by LupusUK »

As I did not have any work to lose I detached/reattached to the project. That got rid of the http file error but it still did not give me any new work

06/01/2012 19:37:09 | Milkyway@home | Detaching from project
06/01/2012 19:37:30 |  | Fetching configuration file from http://milkyway.cs.rpi.edu/milkyway/get ... config.php
06/01/2012 19:37:49 | Milkyway@Home | Master file download succeeded
06/01/2012 19:37:54 | Milkyway@Home | Sending scheduler request: Project initialization.
06/01/2012 19:37:54 | Milkyway@Home | Requesting new tasks for CPU and NVIDIA GPU
06/01/2012 19:37:55 | Milkyway@Home | Scheduler request completed: got 0 new tasks
06/01/2012 19:37:55 | Milkyway@Home | No tasks sent
06/01/2012 19:37:55 | Milkyway@Home | Tasks for CPU are available, but your preferences are set to not accept them
Temujin
Posts: 2259
Joined: Mon Mar 13, 2006 12:00 am

Post by Temujin »

06/01/2012 19:37:49 | Milkyway@Home | Master file download succeeded
From reading the post on the milkyway forum, it looks like you have to press Update untill you see the above message in the messages Tab.
That master file presumably contains new information that corrects the http error.
Just then have to wait for work to be available
Zydor
Posts: 437
Joined: Mon Nov 01, 2010 12:00 am

Post by Zydor »

HTTP errors are invariably (but in very very rare cases are not) a problem on the Server, not the Client end.

What happens (most times) is the server has become flooded with requests for work, work returns etc etc, to the extent it can't cope. In extremis it starts to drop off the sending of acknowledgements that it has received a request which can cause confusion at the Client end. End result is it falls over shuts the door, and says "sod it, no one gets squat until someone sorts out the crowd at the door " .... and refuses to play yelling "http error"

It also frequently occurs during maintenance if the site admin got idle and didnt shut off the world properly before starting maintenance. On those occasions the server says in response to a request for work etc "what you doing here .... sod off", yelling "http service not available"

MW has recently done its move to the new server, it will be getting mega numbers of download requests that could flood it. They were also going to move the maindb elsewhere, that could cause http whilst they do that. Or likely they were doing hot fixes/maintenance "on the fly".

It'll settle over the next few days as they fine tune the server and tidy up.

SETI was a classic for http errors (some say it still is) after they opened up after the weekly maintenance shut down - there was always a massive surge of requests at that point, and http error was pretty well the norm for 2 or 3 hours until the simunltaneous numbers of work requests/uploads/downloads died down. It used to happen a lot at the start of a Prime Grid Challenge as everyone yelled for work at start time plus 0.1 seconds:) Although they worked on that and its much improved now.

It'll settle over the next few days as they fine tune the server and tidy up.

There are other causes of http error at the server end, but 99% of time it just flooded, or a service is not turned on properly due to maintenance, or unusually corruption.

Nothing you can do about it, other than wait it out until the cause of the error is removed at the server end. It has ziltch (99% of the time) to do with your PC, so don't rip it apart when http error occurs as the error is coming from the server, not your PC (99% of the time).

Regards
Zy
UBT - Chris Suddick
Active UBT Contributor 15+ yrs
Posts: 698
Joined: Tue Mar 14, 2006 12:00 am
Location: Warrington, Cheshire

Post by UBT - Chris Suddick »

It would appear that you were all right. I just waited for a few more updates and then it cleared itself and it is now working again. <phew>

Many thanks,

Chris.
Post Reply