Declining tester numbers?
Message boards : Number crunching : Declining tester numbers?
| Author | Message | |
|---|---|---|
|
Seems to me that a percentage of our origional testers are no longer active. This is becoming increasingly evident in the posts about results not able to be validated because there just aren't enough systems downloading work.
|
||
ID:
725 | Rating: 0
| rate:
/
|
||
|
I agree, it is taking over a week for mine to validate. With fewer active users and a minimum quorum of 3, I think we could use some fresh blood. I asked about invitation codes in the Q&A forum 5 days ago and still have not received a reply.
|
||
|
ID:
727 | Rating: 0
| rate:
/
|
||
|
I give you the OK to mail me some, it will not be considered as spam ;)
|
||
|
ID:
728 | Rating: 0
| rate:
/
|
||
Andre Kerstens
Forum moderator Project tester Volunteer tester 
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
That's a good idea. I'll send out the invitation code to all users who have currently more than 0 credits and at least 1 result in our db. Please pass it on to serious alpha testers who don't mind to see lots of errors and problems and maybe even loose their credit if needed.
Seems to me that a percentage of our origional testers are no longer active. This is becoming increasingly evident in the posts about results not able to be validated because there just aren't enough systems downloading work. ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
732 | Rating: 0
| rate:
/
|
||
|
This is a particularly bad problem for intel macs. According to boincstats, there are 208 hosts, and of those, only 12 intel macs. It shows them all as active. But my guess is that it takes some real time before those who have left show as inactive. Also, the message telling mac users to stop crunching is probably contributing to the problem. I think that there are about 3-5 intel mac machines still crunching, and probably a very small percentage of their time. We just can't get WUs out of "pending" status.
|
||
|
ID:
733 | Rating: 0
| rate:
/
|
||
|
I am not sure about declining numbers, my last results were sent to me on the 26/9/06 and I returned them on the 27/09 and 28/09. A number are still pending, I have not received anymore work since the 28/9/06.
|
||
|
ID:
735 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
Although I am not 100% sure if this explanation holds, but I think it has to do with the fact that D@H uses homogeneous redundancy (or HR). For this to work well, we need a good mix of mac, linux and windows machines or else the boinc server shared memory segment will quickly get full of results reserved by windows machines meaning that linux and macs won't get any work (the client tells you that there is work, but that it is committed to other platforms). Since we have so much more windows machines compared to the rest, it is well possible that these steal away all the results in the shared memory segment. Making the shared memory segment bigger will relieve this problem for a short while until again all the results in that bigger memory have been taken by the majority of windows machines again. There is a possible solution for this in the works by the boinc dev team and the world community grid people (who also use HR) in that the shared memory is partitioned, which means every platform will always have some work for them in the queue. The downside is that estimations of the size of these partitions have to be really good or else results might have to wait a long time before being send out (in the case that the partition is too big) or there is not enough work available (in the case the partition is too small).
I am not sure about declining numbers, my last results were sent to me on the 26/9/06 and I returned them on the 27/09 and 28/09. A number are still pending, I have not received anymore work since the 28/9/06. ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
736 | Rating: 0
| rate:
/
|
||
|
waa hoo! i got an invite this evening. Once hashclash it out of work, which shouldnt be long, i can bring a fair amount og computing power here. I currently can do about 6k RAC, but not all of it will go here, for I have 31 other projects i am working on :P.
|
||
|
ID:
742 | Rating: 0
| rate:
/
|
||
|
waa hoo! i got an invite this evening. Once hashclash it out of work, which shouldnt be long, i can bring a fair amount og computing power here. I currently can do about 6k RAC, but not all of it will go here, for I have 31 other projects i am working on :P. Hello Tank_Master, welcome to the project. To give you an idea Linux machines do run the work units quicker than Windows machines due to the nature of the computer programmes making up Docking@home. My linux machine is an AMD Opteron 848 (2 cpus at 2.2 GHz) and each work unit takes 1 hour 27 minutes. From what other testers have said it appears that Windows P4 machines take approx 2 to 3 hours and P3 machine a little bit longer at about 4 hours. I could be wrong on this so are there any Windows testers out there can you give Tank_Master a better esimate? ____________ 
|
||
|
ID:
744 | Rating: 0
| rate:
/
|
||
|
On my Pentium M 2.0 Ghz it took 3.5 hours to complete the first WU |
||
|
ID:
745 | Rating: 0
| rate:
/
|
||
|
So its in par with rosetta and spinhenge, I can manage that. WCG is a bit long, and i can only effectivly run them on my P4s. This sounds like I sould be able to run peoject on a higher end P3 as well. cool. Thx for the info Conan and Christophe Vercammen.(And yes, ill be running windows all off of them, after all, the licences were doneted to us :Þ) |
||
|
ID:
746 | Rating: 0
| rate:
/
|
||
|
Doesn't help that some one has posted the invite code in a public forum, how hard is it to realise that the code is not public and it should be used to invite people along to the projects
|
||
|
ID:
747 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
We'll change the invite code when we reach 180 or so users. Just hope we get more hosts that do NOT run windows, else this exercise was pretty futile :-)
Doesn't help that some one has posted the invite code in a public forum, how hard is it to realise that the code is not public and it should be used to invite people along to the projects ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
749 | Rating: 0
| rate:
/
|
||
|
The code has now been edited out of the post I mentioned before
|
||
|
ID:
750 | Rating: 0
| rate:
/
|
||
|
All workunits on my Kubuntu-Linux makes errors with code 1.
|
||
|
ID:
751 | Rating: 0
| rate:
/
|
||
|
In the main page in the news secction the one posted on September 19, 2006 15:30:00 will help you with a temp fix for your problem. |
||
|
ID:
753 | Rating: 0
| rate:
/
|
||
|
Maybe you should make a download button for Linux application ONLY on the first side of the project and invite anyone to participate. |
||
|
ID:
756 | Rating: 0
| rate:
/
|
||
|
Maybe you should make a download button for Linux application ONLY on the first side of the project and invite anyone to participate. Make that Mac and Linux. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
758 | Rating: 0
| rate:
/
|
||
|
> To help out I added another Linux computer to the list.
|
||
|
ID:
768 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
We've changed the invite code again since we're at 185 users or so. Let's see if that is enough for now. We are aware of the linux/mac 'commited to other platforms' problem and are looking for a suitable solution. Because of the homogeneous redundancy this is not so easy to solve though.
|
||
|
ID:
772 | Rating: 0
| rate:
/
|
||
|
We've changed the invite code again since we're at 185 users or so. Let's see if that is enough for now. Can you please break down that 185 by OS? Specifically, how many Macs? And how many of those reported in the past few days? ____________ Dublin, CA Team SETI.USA |
||
|
ID:
774 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
I'll break it down in platforms:
____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
778 | Rating: 0
| rate:
/
|
||
|
I'll break it down in platforms: Thanks! ____________ Dublin, CA Team SETI.USA |
||
|
ID:
779 | Rating: 0
| rate:
/
|
||
|
Thanks, we were able to recruit 3 or 4 Linux users with the code you provided. I'll be on the lookout for more (and try to find some Mac users) for the next wave. |
||
|
ID:
789 | Rating: 0
| rate:
/
|
||
|
We got at least one PPC Mac user as well, just waiting on the client for it.
|
||
|
ID:
794 | Rating: 0
| rate:
/
|
||
Message boards : Number crunching : Declining tester numbers?
Database Error: The MySQL server is running with the --read-only option so it cannot execute this statement
array(3) {
[0]=>
array(7) {
["file"]=>
string(47) "/boinc/projects/docking/html_v2/inc/db_conn.inc"
["line"]=>
int(97)
["function"]=>
string(8) "do_query"
["class"]=>
string(6) "DbConn"
["object"]=>
object(DbConn)#30 (2) {
["db_conn"]=>
resource(120) of type (mysql link persistent)
["db_name"]=>
string(7) "docking"
}
["type"]=>
string(2) "->"
["args"]=>
array(1) {
[0]=>
&string(50) "update DBNAME.thread set views=views+1 where id=64"
}
}
[1]=>
array(7) {
["file"]=>
string(48) "/boinc/projects/docking/html_v2/inc/forum_db.inc"
["line"]=>
int(60)
["function"]=>
string(6) "update"
["class"]=>
string(6) "DbConn"
["object"]=>
object(DbConn)#30 (2) {
["db_conn"]=>
resource(120) of type (mysql link persistent)
["db_name"]=>
string(7) "docking"
}
["type"]=>
string(2) "->"
["args"]=>
array(3) {
[0]=>
object(BoincThread)#3 (16) {
["id"]=>
string(2) "64"
["forum"]=>
string(1) "2"
["owner"]=>
string(2) "12"
["status"]=>
string(1) "0"
["title"]=>
string(25) "Declining tester numbers?"
["timestamp"]=>
string(10) "1159915050"
["views"]=>
string(4) "1669"
["replies"]=>
string(2) "24"
["activity"]=>
string(20) "2.0394733488579e-129"
["sufferers"]=>
string(1) "0"
["score"]=>
string(1) "0"
["votes"]=>
string(1) "0"
["create_time"]=>
string(10) "1159726595"
["hidden"]=>
string(1) "0"
["sticky"]=>
string(1) "0"
["locked"]=>
string(1) "0"
}
[1]=>
&string(6) "thread"
[2]=>
&string(13) "views=views+1"
}
}
[2]=>
array(7) {
["file"]=>
string(63) "/boinc/projects/docking/html_v2/user/community/forum/thread.php"
["line"]=>
int(184)
["function"]=>
string(6) "update"
["class"]=>
string(11) "BoincThread"
["object"]=>
object(BoincThread)#3 (16) {
["id"]=>
string(2) "64"
["forum"]=>
string(1) "2"
["owner"]=>
string(2) "12"
["status"]=>
string(1) "0"
["title"]=>
string(25) "Declining tester numbers?"
["timestamp"]=>
string(10) "1159915050"
["views"]=>
string(4) "1669"
["replies"]=>
string(2) "24"
["activity"]=>
string(20) "2.0394733488579e-129"
["sufferers"]=>
string(1) "0"
["score"]=>
string(1) "0"
["votes"]=>
string(1) "0"
["create_time"]=>
string(10) "1159726595"
["hidden"]=>
string(1) "0"
["sticky"]=>
string(1) "0"
["locked"]=>
string(1) "0"
}
["type"]=>
string(2) "->"
["args"]=>
array(1) {
[0]=>
&string(13) "views=views+1"
}
}
}
query: update docking.thread set views=views+1 where id=64