Invalid Host Messaging

Message boards : Number crunching : Invalid Host Messaging
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 5 · 6 · 7 · 8 · 9 · 10 · 11 . . . 20 · Next

AuthorMessage
rob smithProject Donor
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 15196
Credit: 251,522,213
RAC: 324,264
United Kingdom
Message 1619459 - Posted: 27 Dec 2014, 20:11:39 UTC

Most of the time its more an annoyance than a problem. It does become a problem when two errant PCs work on the the same work unit and return the same wrong result, which gets incorporated into the project science database.

(A lot of effort is put in by the team members to ensure that from one build to the next the correct data is returned by correctly functioning applications, and as far as possible builds that return errors or incorrect results are consigned to an early bath)
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1619459 · Report as offensive     Reply Quote
Profile JanniCash
Avatar

Send message
Joined: 17 Nov 03
Posts: 57
Credit: 1,276,920
RAC: 0
United States
Message 1619473 - Posted: 27 Dec 2014, 20:58:03 UTC - in response to Message 1619459.  

Most of the time its more an annoyance than a problem. It does become a problem when two errant PCs work on the the same work unit and return the same wrong result, which gets incorporated into the project science database.

Ah, so it is possible that multiple systems produce the same errors rather than random errors, which then results in the erroneous data being considered confirmed.

I had expected that errors, resulting from overclocked/overheated components, are not easily reproducible. But I have to admit that this sort of hardware problems are definitely not my strong side. I know a lot more when it comes to databases and networking issues.

Thanks for your time.
ID: 1619473 · Report as offensive     Reply Quote
Profile David Anderson (not *that* DA)Project Donor
Avatar

Send message
Joined: 5 Dec 09
Posts: 169
Credit: 45,637,680
RAC: 31,473
United States
Message 1619490 - Posted: 27 Dec 2014, 21:53:16 UTC - in response to Message 1619473.  

While I don't know about all results fields, I note that
GPUs being bad often give spike count of 30 so
it appears there is a strong correlation between
the invalid-gpu-results.
Sometimes its various counts sum to 30.

Not random, in other words, but with
a fairly strong pattern.
ID: 1619490 · Report as offensive     Reply Quote
David S
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 18196
Credit: 22,858,068
RAC: 2,961
United States
Message 1627544 - Posted: 14 Jan 2015, 17:25:57 UTC

I just sent a PM to user jedi, host 5925627. It has 221 invalids and 292 inconclusives. Worse, it has 221 valids that are suspect.

http://setiathome.berkeley.edu/workunit.php?wuid=1674186366 is an overflow that matched another overflow and left a good host with an invalid.

http://setiathome.berkeley.edu/workunit.php?wuid=1669848312 is one of those that ran as v6, but somehow managed to validate.

That's as far down his valid list as I went.
David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

ID: 1627544 · Report as offensive     Reply Quote
Profile David Anderson (not *that* DA)Project Donor
Avatar

Send message
Joined: 5 Dec 09
Posts: 169
Credit: 45,637,680
RAC: 31,473
United States
Message 1627559 - Posted: 14 Jan 2015, 18:08:01 UTC

http://setiathome.berkeley.edu/workunit.php?wuid=1672054677
is another in the questionable series, this one is: 19ap11ad.22063.48525.140733193388041.12.99

Most of us processing that wu got invalid, but one task
is now marked inconclusive, so ... that's probably not
a good sign?
ID: 1627559 · Report as offensive     Reply Quote
David S
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 18196
Credit: 22,858,068
RAC: 2,961
United States
Message 1627568 - Posted: 14 Jan 2015, 18:33:17 UTC - in response to Message 1627559.  

http://setiathome.berkeley.edu/workunit.php?wuid=1672054677
is another in the questionable series, this one is: 19ap11ad.22063.48525.140733193388041.12.99

Most of us processing that wu got invalid, but one task
is now marked inconclusive, so ... that's probably not
a good sign?

That's pretty much how all of them look until they finally fill up with invalids. At least in my experience.

But we're getting into the topic of a different thread.
David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

ID: 1627568 · Report as offensive     Reply Quote
WezH
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 571
Credit: 65,752,033
RAC: 2
Finland
Message 1628529 - Posted: 16 Jan 2015, 18:43:37 UTC

Computer 7417673 has problems.

I did send message to it's owner. Wait and see if I receive response.
"Please keep Your signature under four lines so Internet traffic doesn't go up too much"

- In 1992 when I had my first e-mail address -
ID: 1628529 · Report as offensive     Reply Quote
rob smithProject Donor
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 15196
Credit: 251,522,213
RAC: 324,264
United Kingdom
Message 1628621 - Posted: 16 Jan 2015, 21:15:21 UTC

Something very strange with that one - far too many tasks "in progress", no credit, no valid results, all completed tasks actually timed out, some "in progress" tasks with very long deadlines (two months!!).
Looks to be someone who has been trying to "cook the books"....
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1628621 · Report as offensive     Reply Quote
Profile UliProject Donor
Volunteer tester
Avatar

Send message
Joined: 6 Feb 00
Posts: 10848
Credit: 5,743,225
RAC: 0
Germany
Message 1628851 - Posted: 17 Jan 2015, 6:21:29 UTC

ET seems to be trying to phone home, but the user may have suspended the tasks/project and thus they didn't get completed.
Others know more about that.
Richard can you provide a clue, if my thinking is right along those lines.
Pluto will always be a planet to me.

Seti Ambassador
ID: 1628851 · Report as offensive     Reply Quote
WezH
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 571
Credit: 65,752,033
RAC: 2
Finland
Message 1629344 - Posted: 18 Jan 2015, 14:52:24 UTC - in response to Message 1628529.  

Computer 7417673 has problems.

I did send message to it's owner. Wait and see if I receive response.


No response so far, still waiting. Anyone else who want's to send personal message to this user? Maybe some who can speak Italian?

I do have feeling that something is preventing to crunch units, like antivirus software or something else?
"Please keep Your signature under four lines so Internet traffic doesn't go up too much"

- In 1992 when I had my first e-mail address -
ID: 1629344 · Report as offensive     Reply Quote
WezH
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 571
Credit: 65,752,033
RAC: 2
Finland
Message 1633027 - Posted: 26 Jan 2015, 16:25:27 UTC - in response to Message 1629344.  

Computer 7417673 has problems.

I did send message to it's owner. Wait and see if I receive response.


No response so far, still waiting. Anyone else who want's to send personal message to this user? Maybe some who can speak Italian?

I do have feeling that something is preventing to crunch units, like antivirus software or something else?


Something did happen, host hasn't contacted servers since 19th. At least there are no more workunits sent this host, although we have to wait at least to 17th of March to all workunits cleared.
"Please keep Your signature under four lines so Internet traffic doesn't go up too much"

- In 1992 when I had my first e-mail address -
ID: 1633027 · Report as offensive     Reply Quote
Werecow
Avatar

Send message
Joined: 13 Mar 05
Posts: 56
Credit: 2,398,613
RAC: 2,927
United States
Message 1633049 - Posted: 26 Jan 2015, 17:47:22 UTC

Being a relative newcomer, I'm not comfortable with messaging other users about possible problems with -their- systems. This one, though, 6734838, seems to be trashing about 50% of WUs.
ID: 1633049 · Report as offensive     Reply Quote
Profile Bernie VineProject Donor
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 26 May 99
Posts: 9080
Credit: 51,165,354
RAC: 21,223
United Kingdom
Message 1633061 - Posted: 26 Jan 2015, 18:09:25 UTC - in response to Message 1633049.  

Being a relative newcomer, I'm not comfortable with messaging other users about possible problems with -their- systems. This one, though, 6734838, seems to be trashing about 50% of WUs.

Unfortunately you cant message the owner as they are "anonymous"

Nothing that can be done I am afraid
"Proud to be born and bred in Croydon"
ID: 1633061 · Report as offensive     Reply Quote
Tutankhamon
Volunteer tester
Avatar

Send message
Joined: 1 Nov 08
Posts: 6705
Credit: 42,263,760
RAC: 14,543
Sweden
Message 1644610 - Posted: 20 Feb 2015, 19:31:58 UTC
Last modified: 20 Feb 2015, 19:36:07 UTC

Simple (for a programmer) solution to this problem:

Reputation based validation (solves the problem that two rogue computers agree on an overflowed result, which would be entered into the DB as a true result)

No computer with many invalids, or errors, should be allowed to get the canonical result. If both wingmen agree on a result, and they both have bad reputation, the WU should be sent out to another computer, or maybe two computers.

Reputation based granting of new WU's. No computer with lots of invalids or errors should be given more than one task each day. If they fix their crap, the reputation will improve, and they will be given more WU's.

This will not increase the demand on the project system, even if more WU's are sent out to be recrunched. All the thousands upon thousands of WU's the broken systems also trash immediately, will no longer be sent out, due to the Reputation based granting of new WU's.
ID: 1644610 · Report as offensive     Reply Quote
Profile Brent Norman
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 1822
Credit: 106,068,381
RAC: 450,269
Canada
Message 1644719 - Posted: 20 Feb 2015, 23:54:11 UTC - in response to Message 1644610.  
Last modified: 20 Feb 2015, 23:56:53 UTC

another garbage making computer
http://setiathome.berkeley.edu/show_host_detail.php?hostid=7505206

I only counted 4 valid tasks that were compared to a CPU.
ID: 1644719 · Report as offensive     Reply Quote
Werecow
Avatar

Send message
Joined: 13 Mar 05
Posts: 56
Credit: 2,398,613
RAC: 2,927
United States
Message 1650982 - Posted: 9 Mar 2015, 11:39:48 UTC

One beginning to rack up an impressive invalid count:

User : 1677084 (wmagb)
Host : 5754408
GPU/s : NVIDIA GeForce GTX 560 (1024MB) driver: 344.48 OpenCL: 1.1
pm sent :


7378919 doesn't have a noteworthy total, but noticed because this host has been wing man for a number of my inconclusives:

User : 660944 (hellhound)
Host : 7378919
GPU/s : --
pm sent :

No pm sent to either yet; any recommendations I could make to them? I haven't been hanging around the forum long enough to know the specifics of problematic GPU/driver combinations, etc.
ID: 1650982 · Report as offensive     Reply Quote
Werecow
Avatar

Send message
Joined: 13 Mar 05
Posts: 56
Credit: 2,398,613
RAC: 2,927
United States
Message 1651630 - Posted: 11 Mar 2015, 11:35:20 UTC

One host from a veteran user that's had 212 tasks with -185 (0xffffffffffffff47) ERR_RESULT_START exit status since 12 Jan 2015, no valid results:

User : 1882 (rekabis)
Host : 6735185
GPU/s : --
pm sent : 11 Mar 2015
ID: 1651630 · Report as offensive     Reply Quote
NoizBoi

Send message
Joined: 30 Jan 01
Posts: 12
Credit: 16,663,747
RAC: 18,006
United Kingdom
Message 1654544 - Posted: 19 Mar 2015, 14:33:57 UTC

I am getting similar problems - -185 (0xffffffffffffff47) ERR_RESULT_START
This applies to my graphics card CPU (GeForce GT740, clock speed 1084MHz with 4Gb memory)running SETI.

System is:-
Windows 7 Pro 64bit
Intel I5 4690 CPU @3.50GHz 4 Core
Gigabyte GA-Z97X Mobo (black edition)
8Gb Corsair Vengeance memory (2x 4Gb sticks)
120Gb SSD C Drive
2 x 1Tb SATA 3 drives

Windows, BOINC and graphics card are all up to date.
Boinc is default install.

Any ideas anyone?

Could it be something to do with Aero?

System is very clean and tidy (new build in Nov 2014)

Thanks to anyone in advance

Phil.
ID: 1654544 · Report as offensive     Reply Quote
Richard HaselgroveProject Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 11516
Credit: 106,145,562
RAC: 70,034
United Kingdom
Message 1654551 - Posted: 19 Mar 2015, 15:11:23 UTC - in response to Message 1654544.  

I am getting similar problems - -185 (0xffffffffffffff47) ERR_RESULT_START
This applies to my graphics card CPU (GeForce GT740, clock speed 1084MHz with 4Gb memory)running SETI.

Adding the information lower down the page:

couldn't start app: CreateProcess() failed - The process cannot access the file because it is being used by another process.

That sounds possibly like an over-active virus scanner.
ID: 1654551 · Report as offensive     Reply Quote
NoizBoi

Send message
Joined: 30 Jan 01
Posts: 12
Credit: 16,663,747
RAC: 18,006
United Kingdom
Message 1654621 - Posted: 19 Mar 2015, 19:23:36 UTC

Thanks Richard,

I think that it is the Graphics drivers - have rolled back drivers and am now waiting to see results.

There is another post '@PRE-FERMI NVidia GPU users: Important warning' which states that there are issues with NVidia driver versions.

If this fails then I will remove all drivers and software for GPU and re-install the ones on the CD supplied with it when purchased in November as I have had no problems until the last two driver updates this year.

Will keep you updated, although away for two days - will try to respond on Sunday.

Phil.
ID: 1654621 · Report as offensive     Reply Quote
Previous · 1 . . . 5 · 6 · 7 · 8 · 9 · 10 · 11 . . . 20 · Next

Message boards : Number crunching : Invalid Host Messaging


 
©2017 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.