Posts by Ray Still

1) Message boards : Number crunching : CLOSED CLOSED CLOSED (Message 206919)
Posted 8 Dec 2005 by Profile Ray Still
Post:
Thanks, Ageless, for the links. I had the front page, of course, but didn't think to look there. If I had seen the Technical News or Server Status links, I had not noticed them. They were very helpful and informative. I have now bookmarked them for future reference
2) Message boards : Number crunching : Configuring a PORT problem, or firewall (Message 205092)
Posted 6 Dec 2005 by Profile Ray Still
Post:
Every little bit helps. Who is your provider?

The RF Ethernet network beyond your modem works in a somewhat weird way. Unlike DSL it's multiplexed with hundreds of other signals by a device calleed a QAM. outbound packets are done by a device called a QPSK. If you have a poor connection or you are on a noisy node lowering the MTU will decrease the RF footprint, making the packet less vurnerable to static. Depending on how the QAM is configured (QAM 16, QAM 256, QAM 1024) static will destroy a packet.

Your Signal to noise ration (SNR) should be above 35db on the downstream channel and outbound at or above 40DBmV

There should never be a need to lower it below 1400.. If there is your cable company needs to clean up your node.



My provider is Charter Communications (a cable company here in North Alabama). My downstream Signal to Noise Ratio is 36 dB. They don't give the upstream SNR, but they do give the Power Level for down up and downstream. Downstream is 2 dBmV and Up stream is 45 dBmV. I'm using the Configuration Manager on my Motorola Cable modem to retreive these figures. I notice, however, that I have 3 failed uploads and 1 failed download at present, although it is showing one still running. I just lowered the MTU to 576, but it still did not appear to work. I get a temporarily failed message with a timeout error in Messages and a Retry in XX:XX:XX in Transfer. However the one process still shows Running under Work, so I'm changing the MTU back to 1400.
3) Message boards : Number crunching : Configuring a PORT problem, or firewall (Message 203486)
Posted 5 Dec 2005 by Profile Ray Still
Post:
You guys know a lot more about all of this than I do, but I did find something interesting that might be relevant at this page. Near the end of the page it says "If these sizes are too large, continue to lower the MTU sizes until you reach a baseline of 1400..." This is exactly what I had to do to succeed in attaching a project. Don't know if it matters or not, but just wanted to point this out to someone who can understand what they are reading. Although I'm using a cable modem and a router rather than a dialup connection, the page might still apply. (My hardware is an iMac G5 running MacOS 10.4.3 using PPP over a Motorola cable modem and a CompUSA router-although who knows which router was rebranded as CompUSA?)
4) Questions and Answers : Macintosh : Failed to attach new project (Message 202433)
Posted 4 Dec 2005 by Profile Ray Still
Post:
I was wrong. 1472 must not have worked. It has failed since then at 1472. I must have succeeded in changing it to 576 at which point it attached the project. I've tried other number as well 1470 and 1460 did not work, either, although 1400 actually *did* work. Since I have now been able to acquire credits, I was able to post this information here, now. Hope this information will be useful to you. Thanks.
5) Message boards : Number crunching : Configuring a PORT problem, or firewall (Message 202432)
Posted 4 Dec 2005 by Profile Ray Still
Post:
The Mac results:

1472 worked. I have attached a project, although it told me it failed, then when I went to change it to 576, it went ahead and attached it before I could get it changed. Thanks for your help!


So, that pretty much verifies it's not a Windows-only issue.

This should be an IP issue, not a platform issue.

As a general rule though, *nix doesn't set the "don't fragment" flag,, and Windows does.


Slight correction to the message above. I've gotten some credits, so I can post here, now. I must have succeeded in changing it to 576 after all before I got it attached, because I ran into the same problem again later at 1472 and was not able to attach a project at that MTU. However, this time I tried other numbers as well. It failed to attach at 1470 and 1460, however, I succeeded at 1400. Just thought this information might be useful to you. Thanks. (By the way, I'm using a iMac G5 running Mac OS 10.4.3.)
6) Questions and Answers : Macintosh : Failed to attach new project (Message 201834)
Posted 3 Dec 2005 by Profile Ray Still
Post:
I'm just sorry it took so long! Glad you're going now.


Read the thread on NC to which you pointed me. For what the information is worth, I am also using a router, although mine is a CompUSA router (although that may actually be a rebranded version of another).
7) Questions and Answers : Macintosh : Failed to attach new project (Message 201628)
Posted 3 Dec 2005 by Profile Ray Still
Post:
I got instructions from the NC thread, following a link to an _old_ hint, that required lots of Terminal expertise... but then I looked further... I should have known a Mac would have a way to do this without needing the Terminal.

System Preferences, Network, Ethernet, configure manually, MTU is an option. Select "Custom" and try 1472. If that fails, the other thread is recommending 576.

Sigh.


1472 worked. I have attached a project, although it told me it failed, then when I went to change it to 576, it went ahead and attached it before I could get it changed. Thanks for your help!
8) Questions and Answers : Macintosh : Failed to attach new project (Message 201382)
Posted 2 Dec 2005 by Profile Ray Still
Post:
OK. Opened Terminal and did the three pings you suggested. Here are the results. (I'll only post the lines at the beginning and end of the run.)

ping -s 1700 -c 20 setiathome.berkeley.edu
PING setiathome.ssl.berkeley.edu (128.32.18.152): 1700 data bytes

(snipped many lines of data)

--- setiathome.ssl.berkeley.edu ping statistics ---
20 packets transmitted, 20 packets received, 0% packet loss
round-trip min/avg/max/stddev = 102.472/105.500/110.959/2.182 ms

ping -s 1700 -c 20 setiathome.berkeley.edu
PING setiathome.ssl.berkeley.edu (128.32.18.152): 1700 data bytes

(snipped many lines of data)

--- galileo.ssl.berkeley.edu ping statistics ---
20 packets transmitted, 20 packets received, 0% packet loss
round-trip min/avg/max/stddev = 102.850/105.020/108.639/1.355 ms

ping -s 1700 -c 20 setiboincdata.ssl.berkeley.edu
PING setiboincdata.ssl.berkeley.edu (66.28.250.125): 1700 data bytes

(snipped many lines of data)

--- setiboincdata.ssl.berkeley.edu ping statistics ---
20 packets transmitted, 20 packets received, 0% packet loss
round-trip min/avg/max/stddev = 101.106/107.131/159.725/12.503 ms

Just out of curiosity, I typed in ping -s 1700 -60 setiboincdata.ssl.berkeley.edu and received

--- setiboincdata.ssl.berkeley.edu ping statistics ---
60 packets transmitted, 60 packets received, 0% packet loss
round-trip min/avg/max/stddev = 103.319/106.345/117.469/2.702 ms

I can run them again and supply the lines I snipped if you would like. I just snipped them for the sake of brevity.
9) Questions and Answers : Macintosh : Failed to attach new project (Message 200776)
Posted 2 Dec 2005 by Profile Ray Still
Post:
OK, I did a traceroute and then a ping (using Network Utility). Here are the results.

Traceroute has started ...

traceroute: Warning: setiathome.berkeley.edu has multiple addresses; using 128.32.18.152
traceroute to setiathome.ssl.berkeley.edu (128.32.18.152), 64 hops max, 40 byte packets
1 192.168.8.1 (192.168.8.1) 2.233 ms 4.731 ms 0.779 ms
2 * * *
3 * * *
<snipped lines 4-62 which were identical to lines 63 and 64>
63 * * *
64 * * *

****************************************
A PING (10 ping limit) resulted in the following:
****************************************

Ping has started ...

PING setiathome.ssl.berkeley.edu (128.32.18.152): 56 data bytes
64 bytes from 128.32.18.152: icmp_seq=0 ttl=238 time=95.982 ms
64 bytes from 128.32.18.152: icmp_seq=1 ttl=238 time=93.768 ms
64 bytes from 128.32.18.152: icmp_seq=2 ttl=238 time=93.398 ms
64 bytes from 128.32.18.152: icmp_seq=3 ttl=238 time=95.273 ms
64 bytes from 128.32.18.152: icmp_seq=4 ttl=238 time=93.792 ms
64 bytes from 128.32.18.152: icmp_seq=5 ttl=238 time=94.711 ms
64 bytes from 128.32.18.152: icmp_seq=6 ttl=238 time=96.410 ms
64 bytes from 128.32.18.152: icmp_seq=7 ttl=238 time=98.472 ms
64 bytes from 128.32.18.152: icmp_seq=8 ttl=238 time=97.214 ms
64 bytes from 128.32.18.152: icmp_seq=9 ttl=238 time=131.134 ms

--- setiathome.ssl.berkeley.edu ping statistics ---
10 packets transmitted, 10 packets received, 0% packet loss
round-trip min/avg/max/stddev = 93.398/99.015/131.134/10.817 ms


Don't know if this tells you anything useful or not, but there it is.
10) Questions and Answers : Macintosh : Failed to attach new project (Message 199943)
Posted 1 Dec 2005 by Profile Ray Still
Post:
Ray, the first thing I would do is download BOINC V5.2.13. It has a lot of fixes for connection problems, that were located when the "crush" of new users started coming over. It was only released yesterday...

Also, the Number Crunching forum has a lot of info on the connection issue. You can't post there until you have credit, but you can read and see if anyone else having the same problem has gotten it fixed. I suspect that 5.2.13 will solve it for you though. If not, and you don't find something in NC that helps, post back here and I'll drag one of the network guys over to help you out.


OK, I don't see anything in Number Crunching. I downloaded BOINC V5.1.13 and installed it. Failed to attach new project with exactly the same messages. One other thing, I used Select computer to select localhost.
11) Questions and Answers : Macintosh : Failed to attach new project (Message 199471)
Posted 30 Nov 2005 by Profile Ray Still
Post:
Sorry if you've covered this, but I could not find it on the boards.

I've switched from Seti@Home Classic. I activated my account with BOINC, uninstalled Seti@Home Classic, looked in my Library folder for vestiges of classic still left (found none), used Activity monitor to check that there were no Classic processes still running (none), tried resetting the project, logged out of my account on my computer and back in, shut down my computer, waited at least ten minutes and then started it back up.

I recently upgraded my computer to an iMac G5 running Tiger (Mac OS 10.4.3), but I checked Seti@Home Classic on it before switching to BOINC and it was running fine, although I did not try running it as a ScreenSaver. I let it finish the data I was running before turning uninstalling Classic. I am running BOINC Manager 5.2.8. I did not have a password when running Classic before, but I added one when I download BOINC. I have tried using the account key and the password.

It seems to be a problem with the schedulers. I get a message saying that the "Master file download succeeded", however after receiving a message that it is "Requesting 8640 seconds of new work", I receive a message that "Attempting to communicate with [setiboinc.ssl.berkeley.edu] timed out", followed by "Scheduler request to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi failed with a return value of -182", then "No schedulers responded" and "Couldn't connect to URL http://setiathome.berkeley.edu/.Please check URL." Therefore it fails to attach to the project.

Any ideas?






 
©2026 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.