Scheduled Outage of Servers

Questions and Answers : Web site : Scheduled Outage of Servers
Message board moderation

To post messages, you must log in.

AuthorMessage
Graeme

Send message
Joined: 10 Nov 01
Posts: 6
Credit: 2,799,466
RAC: 0
United Kingdom
Message 1026145 - Posted: 18 Aug 2010, 16:00:31 UTC

Given that you have a three day outage of your upload/download servers, is there a way of informing BOINC to back-off communications by a large period of time? It would prevent lots of wasted communication attempts.

Grimble
Registered Linux User #450547
Running KDE 4.5.0 on 2.6.33.5-desktop586-2mnb kernel.
Mandriva Linux release 2010.1 (Official) for i586
ID: 1026145 · Report as offensive
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1026164 - Posted: 18 Aug 2010, 17:09:56 UTC - in response to Message 1026145.  

I believe the BOINC manager allows you to stop the communications on specific days


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1026164 · Report as offensive
Profile Gundolf Jahn

Send message
Joined: 19 Sep 00
Posts: 3184
Credit: 446,358
RAC: 0
Germany
Message 1026186 - Posted: 18 Aug 2010, 18:59:39 UTC - in response to Message 1026164.  

I believe the BOINC manager allows you to stop the communications on specific days

Yes, but it would be easier, especially for people attached to multiple projects, if the scheduler (or whoever) would give the "Project is down for maintenance" message, preferably with a backoff for 24 hours (or so).

Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours
ID: 1026186 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 1026285 - Posted: 19 Aug 2010, 4:02:24 UTC - in response to Message 1026186.  

My BOINC is in project wide back-off on trying to upload its 20 tasks. At the moment still 4 hours 27 minutes to go till the end of that back-off.
ID: 1026285 · Report as offensive
William M. Pierce

Send message
Joined: 26 Sep 05
Posts: 8
Credit: 30,425,862
RAC: 41
United States
Message 1026369 - Posted: 19 Aug 2010, 13:33:15 UTC
Last modified: 19 Aug 2010, 13:33:43 UTC

Are these server outages going to be a regular thing in the future? I have almost exclusively only contributed to the SETI@home project but now all of my pc's and servers are sitting idle because of this downtime and it's happening almost too often for me. Are things going to eventually go back to normal or should this be expected for an extended period of time? Is there anything I can do to help stop these outages?
ID: 1026369 · Report as offensive
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1026380 - Posted: 19 Aug 2010, 14:00:18 UTC - in response to Message 1026369.  

if you are running out of work then its simply a matter of changing your preferences to have more 4 days of work on hand. this keeps you busy during the dry spells


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1026380 · Report as offensive
Graeme

Send message
Joined: 10 Nov 01
Posts: 6
Credit: 2,799,466
RAC: 0
United Kingdom
Message 1026672 - Posted: 20 Aug 2010, 12:22:37 UTC - in response to Message 1026186.  


Yes, but it would be easier, especially for people attached to multiple projects, if the scheduler (or whoever) would give the "Project is down for maintenance" message, preferably with a backoff for 24 hours (or so).

Not just easier, essential if I don't want my other projects affected.
Some figures:
Between Wed18, 2.13 BST and today Fri 20, 10.51, one of my PCs (the one that is on 24/7) made 202 attempts to access the SETI servers. According to BOINC All Project Stats, SETI has over 2,700,000 hosts connected. OK, not all are on 24/7. but if 67% are, that's over 365 million wasted messages.

Grimble
Registered Linux User #450547
Running KDE 4.5.0 on 2.6.33.5-desktop586-2mnb kernel.
Mandriva Linux release 2010.1 (Official) for i586
ID: 1026672 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 1026678 - Posted: 20 Aug 2010, 13:05:10 UTC - in response to Message 1026672.  
Last modified: 20 Aug 2010, 13:07:28 UTC

According to BOINC All Project Stats, SETI has over 2,700,000 hosts connected.

No, according to BOINCStats Seti has over 2.7 million hosts total. Active ones are only 249,187. Which is a quite different number when used further in your calculations.

Now, it's also very unwise to think that just because you 'see' 202 attempts, that all computers out there do so. Here's all the communication attempts and their project wide back-offs that I see on my computer since Tuesday:

17-8-2010 20:18:36 | SETI@home | [file_xfer] project-wide xfer delay for 4418.469802 sec
17-8-2010 21:32:19 | SETI@home | [file_xfer] project-wide xfer delay for 10623.957030 sec
18-8-2010 00:29:46 | SETI@home | [file_xfer] project-wide xfer delay for 28535.441145 sec
18-8-2010 08:25:25 | SETI@home | [file_xfer] project-wide xfer delay for 38691.073336 sec
18-8-2010 19:10:20 | SETI@home | [file_xfer] project-wide xfer delay for 28037.745293 sec
19-8-2010 02:57:41 | SETI@home | [file_xfer] project-wide xfer delay for 26756.260872 sec
19-8-2010 10:23:40 | SETI@home | [file_xfer] project-wide xfer delay for 25307.339702 sec
19-8-2010 17:25:32 | SETI@home | [file_xfer] project-wide xfer delay for 33298.159734 sec
20-8-2010 02:40:33 | SETI@home | [file_xfer] project-wide xfer delay for 32979.766228 sec
20-8-2010 11:50:16 | SETI@home | [file_xfer] project-wide xfer delay for 36544.718772 sec

Just 10!
(The last back-off is happening for all 35 tasks I have ready to upload. One project-wide back-off for all uploads waiting. When the servers will open up later on, it'll try to upload them 2 by 2. Not ALL 35 at the same time!)
ID: 1026678 · Report as offensive
Graeme

Send message
Joined: 10 Nov 01
Posts: 6
Credit: 2,799,466
RAC: 0
United Kingdom
Message 1026737 - Posted: 20 Aug 2010, 17:02:37 UTC - in response to Message 1026678.  

Apologies - I stand corrected. I didn't realise there was an "active hosts" count.
Is it significant that my version of BOINC - 6.10.56 - should try so much harder to connect than yours?
Grimble
ID: 1026737 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 1026742 - Posted: 20 Aug 2010, 17:16:21 UTC - in response to Message 1026737.  

I guess it depends on whether you have a core client configuration file with a line like this in it: <max_file_xfers>X</max_file_xfers>, where X is the amount of transfers you want to have happening at the same time.

But otherwise, no. There's no update on the transfers in 6.11.6, it's working the same way in 6.10.56 and .58.

If all's correct, you're uploading now though. I've only got 3 remaining waiting for upload, the rest are up already since Seti opened the sluice gates.

ID: 1026742 · Report as offensive

Questions and Answers : Web site : Scheduled Outage of Servers


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.