ERROR: Primenet error 2252. Q: which tcp/ip ports are being used for the transfer?
 

$ ./mprime -d
Mersenne number primality test program version 23.5
Getting exponents from server
ERROR: Primenet error 2252
The FAQ at [url]http://www.entropia.com/ips/faq.html[/url] may have more information.
Will try contacting server again in 60 minutes.
Resuming primality test of M18080627 at iteration 18080384 [99.99%]

...

prime.log starts....
[Tue Oct 14 15:51:45 2003 - ver 23.5]
Updating user information on the server
Updating computer information on the server
Sending text message to server:
UID: nevillednz/romeo, UID: nevillednz, User: Neville C. Dempsey, [email]NevilleDNZ.gimps@3ttechnology.com[/email]
Getting exponents from server
Sending expected completion date for M18080627: Nov 30 2003
[Mon Oct 20 12:20:30 2003 - ver 23.5]
Sending text message to server:
UID: nevillednz/romeo, M18080627 completed P-1, B1=210000, B2=4357500, WZ2: D3830A81
Sending result to server for exponent 18080627
[Tue Nov 11 17:14:57 2003 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: Jan 20 2004
[Tue Dec 9 23:31:58 2003 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: Feb 17 2004
[Wed Jan 7 08:18:18 2004 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: Mar 10 2004
[Wed Feb 4 22:06:19 2004 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: Mar 15 2004
[Fri Mar 12 14:25:11 2004 - ver 23.5]
ERROR: Primenet error 2252
[Fri Mar 12 14:25:40 2004 - ver 23.5]
ERROR: Primenet error 2252
[Fri Mar 12 16:40:24 2004 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: Apr 19 2004
[Fri Apr 2 11:26:33 2004 - ver 23.5]
Updating computer information on the server
ERROR: Primenet error 2252
[Fri Apr 2 11:29:44 2004 - ver 23.5]
Updating computer information on the server
ERROR: Primenet error 2252

and continues with error 2252 for months....
except on April 10th!

ERROR: Primenet error 2252
[Sat Apr 10 01:21:11 2004 - ver 23.5]
Updating computer information on the server
ERROR: Primenet error 2252
[Sat Apr 10 02:24:48 2004 - ver 23.5]
Updating computer information on the server
ERROR: Primenet error 2252
[Sat Apr 10 03:27:26 2004 - ver 23.5]
Updating computer information on the server
ERROR: Primenet error 2252
[Sat Apr 10 09:56:27 2004 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: May 9 2004
[Mon Apr 26 18:33:48 2004 - ver 23.5]
Getting exponents from server
ERROR: Primenet error 2252
[Mon Apr 26 19:33:51 2004 - ver 23.5]
ERROR: Primenet error 2252
[Mon Apr 26 20:33:56 2004 - ver 23.5]
ERROR: Primenet error 2252
[Mon Apr 26 21:35:31 2004 - ver 23.5]


I tried with and without a firewall will no success.

Can someone tell me which tcp/ip ports are being used for the transfer? (so I can check the ports are not blocked and the firewall is offline)

NevilleDNZ

Forgive the following verbiage (passwd Xed)
When I run
# ./mprime -c
Sending result to server for exponent 18080627
ERROR: Primenet error 2252
The FAQ at [url]http://www.entropia.com/ips/faq.html[/url] may have more information.


I get the following tcp/http traffic... so it look like maybe I get a dialogue with the mprime server (64.66.6.250)... Any ideas?
# tethereal -i ppp0
Capturing on ppp0
May 1 11:18:36 romeo kernel: device ppp0 entered promiscuous mode
0.000000 220.255.126.175 -> 209.132.176.4 NTP NTP # ignore
0.222821 209.132.176.4 -> 220.255.126.175 NTP NTP # ignore
6.575163 220.255.126.175 -> 64.66.6.250 TCP 49183 > http [SYN] Seq=0 Ack=0 Win=5808 Len=0 MSS=1452 TSV=4083066 TSER=0 WS=0
9.568904 220.255.126.175 -> 64.66.6.250 TCP 49183 > http [SYN] Seq=0 Ack=0 Win=5808 Len=0 MSS=1452 TSV=4083366 TSER=0 WS=0
9.578076 64.66.6.250 -> 220.255.126.175 TCP http > 49183 [SYN, ACK] Seq=0 Ack=1 Win=8190 Len=0 MSS=1412
9.578163 220.255.126.175 -> 64.66.6.250 TCP 49183 > http [ACK] Seq=1 Ack=1 Win=5808 Len=0
9.695874 220.255.126.175 -> 64.66.6.250 HTTP GET /cgi-bin/pnHttp.exe?ar&4&88&4&24506&41925&nevillednz&XXXXXX&romeo&18080627&1&E0DD1ADB296E68AA HTTP/1.0
9.708941 64.66.6.250 -> 220.255.126.175 TCP http > 49183 [RST] Seq=1 Ack=110 Win=9814 Len=0
May 1 11:19:01 romeo kernel: device ppp0 left promiscuous mode

[root@romeo root]# telnet 64.66.6.250 80
Trying 64.66.6.250...
Connected to 64.66.6.250.
Escape character is '^]'.
help
Connection closed by foreign host.

As you figured out mprime contacts 64.66.6.250 over port 80

Any idea what "ERROR: Primenet error 2252" is?
(I appear to have a good connection to primnet and manage to set up a tcp dialogue)

[QUOTE=nevillednz]Any idea what "ERROR: Primenet error 2252" is?
(I appear to have a good connection to primnet and manage to set up a tcp dialogue)[/QUOTE]

Are you using v.23.8 of Prime95? There is some information about "error 2252" on the txt files that come with the software.

Luigi

mprime binary 23.8 for Linux?
 

Thanx, I even found your answer on the main page of [url]http://mersenne.org/[/url]

Under: "Other Recent News

"Error 2252. Some users that have a proxy server between their computer and "mersenne.org are reporting this error. Version 23.8 corrects this problem by using full "URLs instead of relative URLs. Visit the download page to upgrade the client.

The bad news is that I run Linux and there is not version 23.8, does this mean all Linux users are having the same problem? (It would be a shame to throw away my factorising results 99.9% done)
2/10/03 8:31 PM 376628 mprime231.tar.gz
2/20/03 6:07 PM 378035 mprime232.tar.gz
5/1/03 9:30 PM 416006 mprime233.tar.gz
5/24/03 3:51 PM 413203 mprime234.tar.gz
7/3/03 8:34 PM 417531 mprime235.tar.gz

Does that mean I need to download and compile the source and send it in? Is this version 2.38 source?
8/12/03 12:27 AM 2511263 source23.zip

I note that the Windows executable is from 2004, but the source is 2003. Is this a problem?
2/17/04 10:19 PM 1065696 p95v238.exe
2/17/04 10:20 PM 913112 p95v238.zip

Would you like me to send in my Linux binaries if/when I recompile?

ThanX

You can use full URLs in Linux 23.5 by adding the line "UseFullURL=1" to primenet.ini - you'll probably have to create that file.

I created the file, and even put UseFullURL=1 into primenet.ini and prime.ini, saddly no luck.
$ ls -lt *.ini
-rw-rw-r-- 1 nevilled nevilled 215 May 3 09:06 local.ini
-rw-rw-r-- 1 nevilled nevilled 230 May 3 08:32 prime.ini
-rw-rw-r-- 1 nevilled nevilled 13 May 3 08:32 primenet.ini
-rw-rw-r-- 1 nevilled nevilled 0 May 1 10:50 worktodo.ini
$ grep UseFullURL=1 *.ini
prime.ini:UseFullURL=1
primenet.ini:UseFullURL=1
$ ./mprime -d
Mersenne number primality test program version 23.5
Sending result to server for exponent 18080627
ERROR: Primenet error 2252
The FAQ at [url]http://www.entropia.com/ips/faq.html[/url] may have more information.
Will try contacting server again in 60 minutes.
^C
$ strings mprime | egrep -i "primenet.ini|url"
primenet.ini
UseFullURL
$ -- so it seems that the executable does know about primenet.ini & UseFullURL

I note that once a month, near the estimated end date the system manages to communicate a different message with the primenet server. i.e.
[Fri Mar 12 16:40:24 2004 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: Apr 19 2004
[Sat Apr 10 09:56:27 2004 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: May 9 2004

Maybe I should just leave it running for a few more day waiting for the next "end date check in"?? (Or I could change LastEndDatesSent=1081562188 to a smaller number to speed this up?)

I also note that the problem only started on the 12th or March. Prior to this the "client" mprime program did not log any server connection problems/messages (even after several months of running). It is possible that something was changed on the server that triggered this problem? Possibly the mprime "server side" communication service was changed, or maybe you installed a WindowsNT service pack that was patching one of Microsofts famous virus/worm holes? (Which works so well that NO relative URLs work anymore...)

Here is a log when the FIRST problem happened:

Updating computer information on the server
Sending expected completion date for M18080627: Mar 15 2004
[Fri Mar 12 14:25:11 2004 - ver 23.5]
ERROR: Primenet error 2252
[Fri Mar 12 14:25:40 2004 - ver 23.5]
ERROR: Primenet error 2252
[Fri Mar 12 16:40:24 2004 - ver 23.5]
Updating computer information on the server
Sending expected completion date for M18080627: Apr 19 2004
[Fri Apr 2 11:26:33 2004 - ver 23.5]
Updating computer information on the server
ERROR: Primenet error 2252
[Fri Apr 2 11:29:44 2004 - ver 23.5]
Updating computer information on the server
ERROR: Primenet error 2252

(FYI: mprime not been running 100% of the time on this machine)

Any more ideas? (I am leaning towards downloading & compiling the new version for linux)

NevilleD

I've been using Mprime 2.35 on Libranet 2.7, 2.8 and Slackware 9.0 - kernel versions 2.4.20 and 2.6.4 (which I abandoned because it ran slightly slower) - since last September without any connection problems. (Well, OK, occasionally the server may have been unavailable for a couple of hours.) I've used it over external dial up modem and via ADSL. I've run it through a firewall. I've never had to tweak the settings. (So I'm probably not going to be much help :no: )

I take it you can use the box for normal internet access so you know it's not just a dodgy cable or card?

The 23.8 sources are not available for download. We can verify that UseFullURL is working by adding the line Debug=1 to primenet.ini. Then post the debug output from prime.log. I don't think we'll learn anything new, but that's all I can think of.

The only other alternative is to download the sources and build 23.5 with debugging on. If you have any suggested improvements to the primenet.c file I'd love to hear them.

dodgy cable or card?
 

Well... my ISP isnt an angel, but I am on the net right now.. no problems.
If U look at the logs the problem only comes up after a month or so when the Mprime client tries (is due) to check in with the mprime server. And from that failure it tries every hour, and gets a failure.

My thought is maybe it is to do with the phase of the calculation.
Here are the "non error messages" in the file "prime.log"

Updating user information on the server
Updating computer information on the server
Sending text message to server:
UID: nevillednz/romeo, UID: nevillednz, User: Neville C. Dempsey, [email]NevilleDNZ.gimps_spamer@3ttechnology.com[/email]
Getting exponents from server
Sending text message to server:
UID: nevillednz/romeo, M18080627 completed P-1, B1=210000, B2=4357500, WZ2: D3830A81
Sending result to server for exponent 18080627
Updating computer information on the server
Getting exponents from server
Sending result to server for exponent 18080627

Currently I am getting the following error every hour: :sick:
[Mon May 3 09:14:12 2004 - ver 23.5]
Sending result to server for exponent 18080627
ERROR: Primenet error 2252

I have tried running the job in the foreground, background and under cron. :unsure:
Maybe I restired some old data files after my hard disk was fried & died?? :nuke:

BTW: It says 2^18080627-1 isn't prime.... incase anyone wondered! :bounce:

Maybe I can just email my results to someone? :surrender:

NevilleDNZ

My next factoring job.
 

Actually I would not mind starting another factorisation:showoff: , but I guess I need to clear this one first. :cry:

I have turned debug on, and Xed the password out of the result...
$ ./mprime -d
Mersenne number primality test program version 23.5
Sending result to server for exponent 18080627
host = mersenne.org, port = 80
IP-addr = 64.66.6.250
GET [url]http://mersenne.org:80/cgi-bin/pnHttp.exe?ar&4&88&4&20921&7865&nevillednz&XXXXXX&romeo&18080627&1&E0DD1ADB296E68AA[/url] HTTP/1.0

Error in recv call: 104
ERROR: Primenet error 2252
The FAQ at [url]http://www.entropia.com/ips/faq.html[/url] may have more information.
Will try contacting server again in 60 minutes.
^D
$ cat primenet.ini
UseFullURL=1
Debug=1

# su root
# grep 104 /usr/include/asm/errno.h
#define ECONNRESET 104 /* Connection reset by peer */

# ping mersenne.org
PING mersenne.org (64.66.6.250) 56(84) bytes of data.
64 bytes from shawbinary-64-66-6-250.4d.net (64.66.6.250): icmp_seq=1 ttl=114 time=195 ms
64 bytes from shawbinary-64-66-6-250.4d.net (64.66.6.250): icmp_seq=2 ttl=114 time=190 ms
64 bytes from shawbinary-64-66-6-250.4d.net (64.66.6.250): icmp_seq=3 ttl=114 time=190 ms
64 bytes from shawbinary-64-66-6-250.4d.net (64.66.6.250): icmp_seq=4 ttl=114 time=189 ms
64 bytes from shawbinary-64-66-6-250.4d.net (64.66.6.250): icmp_seq=5 ttl=115 time=186 ms

--- mersenne.org ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4064ms
rtt min/avg/max/mdev = 186.513/190.640/195.937/3.099 ms
# telnet mersenne.org 80
Trying 64.66.6.250...
Connected to mersenne.org.
Escape character is '^]'.
123456789
Connection closed by foreign host.
#

I had a chuckle when you said "The 23.8 sources are not available for download". I guess the source is now closed for some good reason... LIN for LEA maybe? :whistle:
([url]http://primes.utm.edu/glossary/page.php?sort=Illegal[/url] ... We would not want to generate an illegal prime by mistake would we :-)

Re: improvements to the mprime source. I would love to make the software a rpm package, with an automatic cron job to restart mprime after system shutdowns. (maybe even a GUI/form for the text input challenged. :alien: )

ECONNRESET - how utterly uninformative. Maybe someone who knows sockets programming far better than I can come up with an explanation for why ping and telnet works, but my simple connect, send, recv does not.

You can email the results to me at [email]woltman@alum.mit.edu[/email] You can get a new exponent from [url]http://mersenne.org/range2.htm[/url]

The only reason 23.8 sources aren't available is I've been too lazy to put them together and upload them. The sources that are there are nearly identical to the 23.8 sources - just a few bug fixes since then.

Good news???
 

I manually entered the URL (produced by debug) into Galeon (A modern/nice web browser): :devil:
[url]http://mersenne.org/cgi-bin/pnHttp.exe?ar&4&88&4&20921&7865&nevillednz&XXXXXX&romeo&18080627&1&E0DD1ADB296E68AA[/url]

And got as a reply:
pnResult=0

It seems like a very small result for such alot of CPU cycles. I hope there is some kind of parity/CRC check in it.

Then I checked my primenet_report:
[url]http://www.mersenne.org/cgi-bin/primenet_report.pl[/url]

And got:
------- Exponents Cleared since last Synchronization -------

prime fact Lucas-Lehmer residue or factor
exponent bits [residues partially masked] date returned computer ID
-------- ---- -- -------------------------------- --------------- ------------
18080627 67 0xE0DD1ADB296E68__ 03-May-04 23:42 romeo

I think this is good news???

Also:
Mersenne PrimeNet Server 4.0 (Build 4.0.031)
Cleared Exponents Report 04 May 2004 00:00 (May 3 2004 5:00PM Pacific)
prime fact Lucas-Lehmer residue or factor
exponent bits [residues partially masked] date returned account ID computer ID
-------- ---- -- -------------------------------- --------------- -------------- ------------
18080627 67 0xE0DD1ADB296E68__ 03-May-04 23:42 nevillednz romeo

Also:
Mersenne PrimeNet Server 4.0 (Build 4.0.031)
Assigned Exponents Report 04 May 2004 00:00 (May 3 2004 5:00PM Pacific)
prime fact current days
exponent bits iteration run / to go / exp date updated date assigned account ID computer ID
-------- -- ---- --------- ----------------- --------------- --------------- -------------- ------------

18079903 66 76.8 -54.8 5.2 18-Feb-04 05:08 17-Feb-04 06:17 dado55 yoo
18080747 67 288522 278.5 6.3 66.3 01-May-04 07:31 30-Jul-03 12:59 S118086 C5C77BD48
18081691 67 2264485 49.8 -9.7 50.3 12-Apr-04 08:55 15-Mar-04 06:21 S154512 C295725BE

romeo with 18080627 nolonger appears. So... I think I am done.

It seems that the URL generated by mprime can be entered directly into any other web browser to affect the mprime server update. I this is so, then this means the problem is in the mprime client software. Maybe something as simple as a carriage return being sent when a linefeed is expected??? When I do my next job I will try the static linked version of mprime. This eliminates other permutations as we can be sure that some local library call isn't messing around with the URL stream.

posted by: Pablo the Duck's Avatar
> I've been using Mprime 2.35 on Libranet 2.7, 2.8 and Slackware 9.0 - kernel versions
> 2.4.20 and 2.6.4 (which I abandoned because it ran slightly slower) - since last
> September without any connection problems.

In case other linux clients encounter a similar problem with connecting to mprime server:
I am using Mprime 2.35 on Redhat9 kernel 2.4.20-28.9.
And rpm glibc-2.3.2-27.9.7 for /lib/ld-linux.so.2 libm.so.6 libc.so.6 (as demanded by mprime client)

Im done. :coffee: Time for my next job...

Thanx
NevilleD

Further (good) news....
the problem mysteriously "went away" on the 12th of May.

So I am extra happy...

i.e. problem WAS:

Error in recv call: 104
ERROR: Primenet error 2252
The FAQ at [url]http://www.entropia.com/ips/faq.html[/url] may have more information.
Will try contacting server again in 60 minutes.
[May 12 07:05] Iteration: 660000 / 24362557 [2.70%]. Per iteration time: 0.335
sec.
Contacting PrimeNet Server.
host = mersenne.org, port = 80
IP-addr = 64.66.6.250
GET [url]http://mersenne.org:80/cgi-bin/pnHttp.exe?ps&32516&.&[/url]. HTTP/1.0

Error in recv call: 104
ERROR: Primenet error 2252
The FAQ at [url]http://www.entropia.com/ips/faq.html[/url] may have more information.
Will try contacting server again in 60 minutes.
[May 12 08:01] Iteration: 670000 / 24362557 [2.75%]. Per iteration time: 0.335
sec.


THEN after the login session was killed and restarted:

Mersenne number primality test program version 23.5
Resuming primality test of M24362557 at iteration 810113 [3.32%]
[May 13 09:43] Iteration: 820000 / 24362557 [3.36%]. Per iteration time: 0.367
sec.
[May 13 10:44] Iteration: 830000 / 24362557 [3.40%]. Per iteration time: 0.367
sec.
[May 13 11:45] Iteration: 840000 / 24362557 [3.44%]. Per iteration time: 0.367
sec.
[May 13 12:47] Iteration: 850000 / 24362557 [3.48%]. Per iteration time: 0.368
sec.
[May 13 13:48] Iteration: 860000 / 24362557 [3.53%]. Per iteration time: 0.367
sec.

"I didn't change anything".... Here are the RPMs I installed around that date (May) for reference incase someone else has a similar problem...
Sat May 08 2004 ethereal-gnome
Sat May 08 2004 lynx
Sat May 08 2004 perl-CGI
Sun May 09 2004 lha
Sun May 09 2004 libpng
Sun May 09 2004 libpng10
Sun May 09 2004 openoffice
Sun May 09 2004 openoffice-i18n
Sun May 09 2004 openoffice-libs
Sun May 09 2004 utempter
Sun May 09 2004 xchat

/var/log/messages.3:May 1 08:03:04 romeo syslogd 1.4.1: restart.
/var/log/messages.3:May 1 21:51:00 romeo syslogd 1.4.1: restart.
/var/log/messages.2:May 2 04:02:38 romeo syslogd 1.4.1: restart.
/var/log/messages.2:May 7 08:30:17 romeo syslogd 1.4.1: restart.
/var/log/messages.1:May 9 04:02:11 romeo syslogd 1.4.1: restart.
/var/log/messages.1:May 9 18:47:24 romeo syslogd 1.4.1: restart.
/var/log/messages.1:May 12 21:29:36 romeo syslogd 1.4.1: restart. <= maybe this reboot loaded an updated library.
/var/log/messages.1:May 12 21:35:34 romeo syslogd 1.4.1: restart.
/var/log/messages.1:May 13 07:27:57 romeo syslogd 1.4.1: restart.

The only thing that vaugely looks like a network/html RPM was lynx (or ethereal).... This was previously not installed, and I am reasonably sure it is not required for mprime.

Venimus, Vidimus, Dolvimus
NevilleDNZ