Pause in work-units
log in

Advanced search

Message boards : News : Pause in work-units

1 · 2 · Next
Author Message
Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 496
Credit: 13,285,798
RAC: 1,758
Message 3528 - Posted: 4 Nov 2014, 2:54:02 UTC

We'll be halting the WUs shortly whilst I upgrade the servers and the software.
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

Jozef J
Send message
Joined: 12 Oct 13
Posts: 18
Credit: 36,095,318
RAC: 28,951
Message 3533 - Posted: 8 Nov 2014, 14:42:20 UTC

So you started today? Because I no longer receive tasks for computers ... ???

Balco Bulls
Send message
Joined: 3 Apr 14
Posts: 3
Credit: 2,682,850
RAC: 752
Message 3534 - Posted: 9 Nov 2014, 5:45:38 UTC - in response to Message 3528.

Is there a time frame we should expect with this upgrade? Just making sure my system will be up and running during that time.

Profile Devlin85
Send message
Joined: 2 May 14
Posts: 1
Credit: 6,110,245
RAC: 1,039
Message 3535 - Posted: 9 Nov 2014, 6:26:27 UTC

noooooooooooo!!! ya i ran out of tasks yesterday/today.. please be a good upgrade! :)

Profile Elektra*
Avatar
Send message
Joined: 12 May 14
Posts: 127
Credit: 8,060,292
RAC: 0
Message 3536 - Posted: 9 Nov 2014, 8:44:56 UTC

Come on Kevin! In 2 days and 18 hours it's time to fill the bunkers for the " Edwin Powell Hubble - 125 anniversary of the birth!"-challenge at Boincstats. Why must you admins always fiddle around with your servers just before challenge starts? :P
____________

Profile alexbeckley
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Jan 13
Posts: 345
Credit: 5,696,421
RAC: 0
Message 3537 - Posted: 10 Nov 2014, 0:05:42 UTC - in response to Message 3536.

Unfortunately then and when and the how of server maintenance is not always completely in our control. But we are endeavouring to have the servers up and running again as soon as possible.
____________
Cheers
Alex
theSkyNet.org webMaster
ICRAR

Kirsten
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 12 Sep 13
Posts: 48
Credit: 4,350,677
RAC: 0
Message 3540 - Posted: 11 Nov 2014, 8:06:26 UTC

As Alex said - we can't always make these things happen when we want them. Kevin let us know he's still waiting on a few galaxy results to come back before he can unplug and upgrade the disks, so it's a waiting game that we're not able to know the end of at the moment - so sorry guys that your prep for the challenge won't go as easily as usual!
____________
Member of theSkyNet team.

Profile Elektra*
Avatar
Send message
Joined: 12 May 14
Posts: 127
Credit: 8,060,292
RAC: 0
Message 3541 - Posted: 11 Nov 2014, 16:24:44 UTC - in response to Message 3540.
Last modified: 11 Nov 2014, 16:44:13 UTC

11.11.2014 16:23 UCT
Tasks ready to send 2
Tasks in progress 22,977

These few galaxy results Kevin is waiting for aren't the 23,000 tasks in progress, are they? If start time of the upgrade depends on all tasks in progress being finished and validated, we'll accept this, of course, but we'd like to know so that the challenge at Boincstats might be postponed until the update work is done and the servers are running stable again.

And of course thank you for your work and messages and good luck for the outstanding maintenance work :)

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 496
Credit: 13,285,798
RAC: 1,758
Message 3546 - Posted: 13 Nov 2014, 4:52:10 UTC

Unfortunately I need most of the WUs back before I can upgrade the disks to bigger ones. The new disks are attached, ready to go. Once I'm down to a few thousand - I'll do the switch.

We only have another 4,000 galaxies ready at the moment, but we'll be queuing up a lot more once David gets back to the USA (he has been here in Australia with me planning the new major release of POGS).

I also need to optimise the database and that is best done when I have most of the pixels back
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

Snow Crash
Send message
Joined: 8 Sep 12
Posts: 10
Credit: 5,024,694
RAC: 0
Message 3550 - Posted: 15 Nov 2014, 12:23:38 UTC - in response to Message 3546.
Last modified: 15 Nov 2014, 12:24:57 UTC

Best of luck on the upgrade ... I know "luck" has nothing to do with it (I do sftw dev) but sometimes it really does feel that way !!!

...[David] has been here in Australia with me planning the new major release of POGS.

Can you share some of the plan?

<humor cap required>
OK, so you'll tell us when you're good and ready ... can you tell us when you will be able to tell us? :-)
That's the problem with this internet stuff, someone is always trying to read something that you didn't actually write, the inevitable whining starts when you won't then clarify something you never said :-)
</humor cap>

I wish the project all the best, I'll be back crunching here soon - cheers!

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 496
Credit: 13,285,798
RAC: 1,758
Message 3553 - Posted: 17 Nov 2014, 4:48:26 UTC

Sorry folks - I'm having some real issues with the permissions on files.

I've copied the data over 5 times now and BOINC doesn't like the permissions for some reason and won't start properly. Hopefully the BOINC Guru's at Berkeley will be able to help. They've offered to debug the thing for me.
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

Profile Presrvd
Avatar
Send message
Joined: 10 Sep 14
Posts: 1
Credit: 14,471,029
RAC: 18,846
Message 3554 - Posted: 17 Nov 2014, 16:23:05 UTC - in response to Message 3553.

Sometimes, it could be more about HOW you copied the files, which I've learned the hard way in the past. A lot of DB systems don't like the standard Windows Copy+Paste solution. I utilize robocopy with a /MIR to retain all file attributes when moving critical files, and rarely run into issues. That is, of course, assuming (big assumption) that you are even running Windows....or that you haven't already tried robocopy. Just a little insight, not trying to be pretentious, or tell you you're doing anything incorrectly.

Berserker
Send message
Joined: 23 Sep 13
Posts: 6
Credit: 21,254,526
RAC: 1
Message 3555 - Posted: 17 Nov 2014, 16:59:30 UTC - in response to Message 3554.

assuming (big assumption) that you are even running Windows

The BOINC server is almost certainly running Linux (I don't know of a BOINC server that runs on Windows - I don't even know if it can).

Often this sort of thing will turn out to be something obvious that frustrates the original developer for hours - or days - upon end (been there!) but is quickly spotted by a fresh pair of eyes, so hopefully the Berkeley folks can get to the bottom of it in short order. It's just annoying when the not-so-obvious obvious solution is found!

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 496
Credit: 13,285,798
RAC: 1,758
Message 3556 - Posted: 18 Nov 2014, 1:18:28 UTC
Last modified: 18 Nov 2014, 1:18:52 UTC

New disks are in and running (a hard link had failed in the linux filesystem and wasn't being copied)

Next task compile the new version of BOINC (after I've taken a snapshot so I can back it out if needed)
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

bassiexp
Send message
Joined: 1 Apr 14
Posts: 3
Credit: 1,933,636
RAC: 0
Message 3557 - Posted: 18 Nov 2014, 2:10:03 UTC - in response to Message 3556.
Last modified: 18 Nov 2014, 2:10:21 UTC

Just curious,

Wasn't the backend on amazon AWS. Saw that from linux conference early januari

This is the one:
http://d0.awsstatic.com/architecture-diagrams/customers/ICRAR-arch-diag.png

Or are these virtual disk spaces.

Profile alexbeckley
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Jan 13
Posts: 345
Credit: 5,696,421
RAC: 0
Message 3558 - Posted: 18 Nov 2014, 4:55:45 UTC - in response to Message 3557.

Well spotted bassiexp
The backend is hosted on AWS, and yes the disks that are being 'swapped out' are virtual ones. Much of the same server administration stuff still applies when using AWS hosts.
____________
Cheers
Alex
theSkyNet.org webMaster
ICRAR

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 496
Credit: 13,285,798
RAC: 1,758
Message 3559 - Posted: 18 Nov 2014, 9:21:29 UTC

Database optimised
Nearly there - downloaded the latest BOINC code - tried to make it and it failed.
Now trying to figure out what has changed...
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

Profile jay
Send message
Joined: 6 May 14
Posts: 3
Credit: 543,419
RAC: 116
Message 3560 - Posted: 18 Nov 2014, 21:22:18 UTC

Hey Kevin,

Thanks for the updates!

Wish I could buy you a beer. (after work is dine. :-) )

Jay

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 496
Credit: 13,285,798
RAC: 1,758
Message 3561 - Posted: 19 Nov 2014, 0:45:25 UTC - in response to Message 3560.

@jay

Thanks
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

Profile Kevin
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar
Send message
Joined: 27 Jul 12
Posts: 496
Credit: 13,285,798
RAC: 1,758
Message 3562 - Posted: 19 Nov 2014, 0:56:45 UTC

The new BOINC code is installed and running.

We're now just waiting for the next batch of Galaxies
____________
Regards
Kevin
-----
International Centre for Radio Astronomy Research

1 · 2 · Next

Message boards : News : Pause in work-units


Main page · Your account · Message boards


Copyright © 2016 The International Centre for Radio Astronomy Research