Successful Lab Repairs


log in

Advanced search

Message boards : News : Successful Lab Repairs

Previous · 1 · 2 · 3
Author Message
Gary
Send message
Joined: 9 Jan 13
Posts: 6
Credit: 537,346
RAC: 1,071
United Kingdom
Message 1344089 - Posted: 8 Mar 2013, 17:42:15 UTC - in response to Message 1343189.

A temporary fix will be good for me, i wont be spending £20 a month on wasted electricity!

How do I 'use a proxy'?

Many thanks,

Gary

Gary
Send message
Joined: 9 Jan 13
Posts: 6
Credit: 537,346
RAC: 1,071
United Kingdom
Message 1344098 - Posted: 8 Mar 2013, 18:06:27 UTC - in response to Message 1343189.


Add to that fact that those having issues, if they use a proxy, will then get excelent download rates, till the proxy gets closed down.


How do i use a proxy?

Gary
Send message
Joined: 9 Jan 13
Posts: 6
Credit: 537,346
RAC: 1,071
United Kingdom
Message 1344108 - Posted: 8 Mar 2013, 18:18:24 UTC - in response to Message 1341700.

Interesting, it never seemed to me that there was a bandwidth problem. To me donwloading SAH takes seconds, AP minutes.

Except in the hours directly after a maintainance outage that is. So I thought that getting the higher bandwidth would primarily shorten the time until the project is back to normal operation after an outage at the current "crunching rate" (which will obviously increase due to Moores law).

But I'm not sure how fast the splitters are, so increasing bandwitdh could have a much lower impact than thought by many if splitters are (or will become at the higher bandwidth) the bottleneck.

Maybe you just discovered one of the extremely rare cases in which neither your bandwidth nor the download servers' bandwidth is the data bottleneck but rather some provider network that your packets pass between you and UC Berkeley? You could test it by downloading anything else from any Berkeley site.



I just downloaded a couple of pdf files from the berkeley site and I got 500kbps+ so that would rule out problems between where I am located and the lab, that is unless the university hosts its website on a server in my back yard!

Since my initial message about slow speed I have had the Seti@home project suspended in my BOINC manager, 9 days of just downloading later and I have only around 2 days worth of work units to process.

Its ok if you are funded for this sort of research but I am not and adding £240 per year to my energy bill (thats around $360) for absolutely no reason is not good.

At the rate I am downloading I could leave my computer on for a full year and it would only be processing data for 1 month out of 12!

Is there no scope for doubling the bandwidth at the lab (upstream, doubling your downstream would do nothing for us!)??

It is worth noting that when I complete a work unit it takes only a few seconds to upload to the lab, this would also suggest that there is no problem with the long connection between the UK and California.

Is there any direct way I can connect to a server and download work??

really, anything is better than this.

Profile Gary CharpentierProject donor
Volunteer tester
Avatar
Send message
Joined: 25 Dec 00
Posts: 12976
Credit: 7,661,111
RAC: 9,236
United States
Message 1344112 - Posted: 8 Mar 2013, 18:40:42 UTC - in response to Message 1344108.

I just downloaded a couple of pdf files from the berkeley site and I got 500kbps+ so that would rule out problems between where I am located and the lab, that is unless the university hosts its website on a server in my back yard!

Not necessarily. UC Berkeley operates multiple interfaces facing CalRen and those have backbone speeds of 10 Gigs/sec where as the link to the SSL only supports 100 Meg/Sec or 1% of that speed. Also the paths packets take to UCB via CalRen will be a different route than the packets to the SSL which is routed over HE.

____________

Gary
Send message
Joined: 9 Jan 13
Posts: 6
Credit: 537,346
RAC: 1,071
United Kingdom
Message 1344658 - Posted: 9 Mar 2013, 17:51:35 UTC - in response to Message 1344112.

I just downloaded a couple of pdf files from the berkeley site and I got 500kbps+ so that would rule out problems between where I am located and the lab, that is unless the university hosts its website on a server in my back yard!

Not necessarily. UC Berkeley operates multiple interfaces facing CalRen and those have backbone speeds of 10 Gigs/sec where as the link to the SSL only supports 100 Meg/Sec or 1% of that speed. Also the paths packets take to UCB via CalRen will be a different route than the packets to the SSL which is routed over HE.


All of the TCP stuff looks a bit complicated. I am not comfortable adjusting my registry for Seti@home, as good a project as it is.

Bill Barto
Send message
Joined: 28 Jun 99
Posts: 860
Credit: 22,235,585
RAC: 10,942
United States
Message 1344660 - Posted: 9 Mar 2013, 18:11:56 UTC

Have you looked at the Windows TCP Settings thread in the Number Crunching forum yet? There is a lot of helpfull info there. I did it and have no troubles downloading now. It's slow but I have no backlog of downloads now. This is a news thread. It is not for discussing problems.

Gary
Send message
Joined: 9 Jan 13
Posts: 6
Credit: 537,346
RAC: 1,071
United Kingdom
Message 1344935 - Posted: 10 Mar 2013, 12:54:42 UTC - in response to Message 1344660.

Have you looked at the Windows TCP Settings thread in the Number Crunching forum yet? There is a lot of helpfull info there. I did it and have no troubles downloading now. It's slow but I have no backlog of downloads now. This is a news thread. It is not for discussing problems.



Strangely, my computer now has 10 days worth of work, no pending downloads and has processed 26 units since yesterday.

It would appear that for now at least the problem has sorted itself out. I have not done anything to my TCP settings either.

Many thanks,

Gary

WinterKnight
Volunteer tester
Send message
Joined: 18 May 99
Posts: 8737
Credit: 25,596,728
RAC: 12,408
United Kingdom
Message 1344947 - Posted: 10 Mar 2013, 14:22:07 UTC - in response to Message 1344935.

Have you looked at the Windows TCP Settings thread in the Number Crunching forum yet? There is a lot of helpfull info there. I did it and have no troubles downloading now. It's slow but I have no backlog of downloads now. This is a news thread. It is not for discussing problems.



Strangely, my computer now has 10 days worth of work, no pending downloads and has processed 26 units since yesterday.

It would appear that for now at least the problem has sorted itself out. I have not done anything to my TCP settings either.

Many thanks,

Gary

That is probably because the splitters have completed all the AP tasks, as displayed on the server status page. So that for the moment only Enhanced multibeam tasks are being downloaded.

ClaggyProject donor
Volunteer tester
Send message
Joined: 5 Jul 99
Posts: 4208
Credit: 34,466,880
RAC: 19,655
United Kingdom
Message 1344959 - Posted: 10 Mar 2013, 15:25:03 UTC - in response to Message 1344947.

Have you looked at the Windows TCP Settings thread in the Number Crunching forum yet? There is a lot of helpfull info there. I did it and have no troubles downloading now. It's slow but I have no backlog of downloads now. This is a news thread. It is not for discussing problems.



Strangely, my computer now has 10 days worth of work, no pending downloads and has processed 26 units since yesterday.

It would appear that for now at least the problem has sorted itself out. I have not done anything to my TCP settings either.

Many thanks,

Gary

That is probably because the splitters have completed all the AP tasks, as displayed on the server status page. So that for the moment only Enhanced multibeam tasks are being downloaded.

Also we aren't overwhelmed with Shorties eithier,

Claggy

Profile Chris SProject donor
Volunteer tester
Avatar
Send message
Joined: 19 Nov 00
Posts: 32306
Credit: 14,270,756
RAC: 11,094
United Kingdom
Message 1344964 - Posted: 10 Mar 2013, 15:33:46 UTC

Also we aren't overwhelmed with Shorties eithier,


Actually I am! Most take 4 hours, I have a load of 1h 30m ones ....

Previous · 1 · 2 · 3

Message boards : News : Successful Lab Repairs

Copyright © 2014 University of California