[QUOTE=pinhodecarlos;432336]Create multiple folders for ecmclient.[/QUOTE]
I was afraid of that. Guess that's what bash scripts are for. Turns out that the server doesn't like 40 clients hitting it at once. :smile:

[CODE]for i in `seq 1 40`; do
cd core${i}
./ecmclient &
cd ..
sleep 1
done
[/CODE]

[QUOTE=frmky;432348]I was afraid of that. Guess that's what bash scripts are for. Turns out that the server doesn't like 40 clients hitting it at once. :smile:

[CODE]for i in `seq 1 40`; do
cd core${i}
./ecmclient &
cd ..
sleep 1
done
[/CODE][/QUOTE]I'm not surprised...

If you could keep the contact interval measured in hours rather than minutes, that would also help.

Paul

Left it at the default, 12 hours I think...

[QUOTE=pinhodecarlos;432336][QUOTE=frmky;432335]Thanks. It doesn't seem to have an option to launch x instances of ecm. How do you deal with multiple cores? One ecmclient/core? If so, can they all run from the same directory? Thanks![/QUOTE]

Create multiple folders for ecmclient.[/QUOTE]

I made some source changes a number of years ago to allow multiple copies of the client to run out of the same directory. See [URL="http://mersenneforum.org/showthread.php?p=232237#post232237"]this[/URL] for details on how it works.

I can send you my mods if you want.

[QUOTE=frmky;432351]Left it at the default, 12 hours I think...[/QUOTE]

I might recommend something a bit shorter than that. The server will hand out the same composite to multiple clients, and 12 hours is probably enough time (depending on how fast your machines are) to significantly overflow the desired curve counts at the current B1 level. I'm using 1 hour right now.

[QUOTE=jyb;432361]I might recommend something a bit shorter than that. The server will hand out the same composite to multiple clients, and 12 hours is probably enough time (depending on how fast your machines are) to significantly overflow the desired curve counts at the current B1 level. I'm using 1 hour right now.[/QUOTE]

Agree with you. I've set in first moment to 6 hours but then after I found twice the same factor in different cores I decided to reduced client-server communication to one hour.

[QUOTE=jyb;432361]I might recommend something a bit shorter than that. The server will hand out the same composite to multiple clients, and 12 hours is probably enough time (depending on how fast your machines are) to significantly overflow the desired curve counts at the current B1 level. I'm using 1 hour right now.[/QUOTE]Once per hour may be ok, but Greg has 40 clients so they'll average once per minute or so. Should still be fine but if hundreds, or thousands, of clients show up I may be forced to review.

I'm very tempted to re-disable the HTML output. The server is showing signs of distress.

Paul

I'll back mine off to once every 3 hours.

Update
 

The updates tables will hit the web pages shortly. There have been so many factors found in April, 275 to be precise, that I won't list them all here. There are now 1544 composites remaining.


Paul

Morning Paul. Can't reach ecmserver. Thank you in advance.

[QUOTE=pinhodecarlos;434620]Morning Paul. Can't reach ecmserver. Thank you in advance.[/QUOTE]Nothing obviously wrong with it;
[code]
[2016-05-22 18:44:31 GMT] Total Time: 55:18:11 Curves: 3594 Total Work: 17246412 Total Factors Found: 1
[2016-05-22 18:44:31 GMT] Connection from 192.168.1.1
[2016-05-22 18:44:31 GMT] pcl@horus (horus9) at 192.168.1.1: Sent GC_7_996 : 3734 ECM curves with B1=11000000
[/code]
One of my machines found a factor earlier today:
[code]
[2016-05-22 08:00:15 GMT] GC_8_966: prime factor returned by pcl@horus (horus9)! Factor=6760281230040088923139812900039408849091673 Method=ECM B1=11000000 Sigma=1
[2016-05-22 08:00:15 GMT] GC_8_966: Composite factor returned by pcl@horus (horus9)! Factor=386278789977361948390438783805711686965369076531572242756537707821327413506724148565973594416891905119724513613634913785778597830321612061459111323763052361385648823514577983077857489214850204851161254009985604123969530665930948311234240504230172151950229657953226407476163820535539820765369313122726963144172588809780535986099147061042847109858112230430810171542946948390884285090566010108035859650749499172113530210358022709009758300467104198873259228009957016895217204532478024213262115311825010535693307154043387264745312389192006856166739432112431164191099576077632378288794963928634227221104702125999590262680426527452456096137858443179993126469310599909270158048290503250126870657668465981955191988760522843449236506633146318680608327524044105534571 Method=ECM B1=11000000 Sigma=1
[/code]
Did it fix itself?