Linux mining distro for the Raspberry PI - MinePeon

[LogicalUnit]

Need help: cgminer 3.3.4 on MinePeon 0.2.3a restarts every few hours. Can someone tell me how to look at the log files?

[1714144198]

1714144198

[1714144198]

1714144198

[Scorpia]

Need help: cgminer 3.3.4 on MinePeon 0.2.3a restarts every few hours. Can someone tell me how to look at the log files?

im not sure about the log files sorry but i would suggest you either go back to 0.2.2 or update minepeon via the git site.

off the top of my head you need to

cd /opt/minepeon
git pull

then optionally

cd http
git pull

then reboot

note this will get you the current unfinished interface but most are finding it more stable than the standard setup on 0.2.3 but the webinterface still has some bugs. which is why i said the http stuff is optional.

ps, more info over at http://www.minepeon.com/forums/index.php

[tk1337]

Need help: cgminer 3.3.4 on MinePeon 0.2.3a restarts every few hours. Can someone tell me how to look at the log files?

Your typical linux log files are in /var/log/

to read the log files within you will need to use sudo

Examples:

Code:

sudo tail errors.log 
sudo cat messages.log

Going off of the dates of the files from within /var/log/, it looks like the log files that are being used by the system are:
auth.log
crond.log
ddemon.log
errors.log
everything.log
faillog
kernel.log
lastlog
messages.log

So if you know what you're looking for in regards to linux OS, then it should be simple for you to find it.

[Grouver (BtcBalance)]

I am running two BFL Singles at the moment with Minepeon. I am getting 58% hw errrors. This isnt right is it?

[tk1337]

I am running two BFL Singles at the moment with Minepeon. I am getting 58% hw errrors. This isnt right is it?

surely doesn't seem right...

if I had access to a BFL, I'd know more to help ya, sorry.

[Beastlymac]

I am running two BFL Singles at the moment with Minepeon. I am getting 58% hw errrors. This isnt right is it?

what version of minepeon are you using? also try bfgminer instead of cgminer or just update cgminer.

[Grouver (BtcBalance)]

what version of minepeon are you using? also try bfgminer instead of cgminer or just update cgminer.

1) Version 0.2.3
2) Tried that. BFGminer says 2%~ But the minepeon backend says 58%~ (edit: now 87% Some engines even say they get 97.5 % hardware errors)
3) my cgminer is already fully updated since the latest version came with Minepeon 0.2.3

[Scorpia]

what version of minepeon are you using? also try bfgminer instead of cgminer or just update cgminer.

1) Version 0.2.3
2) Tried that. BFGminer says 2%~ But the minepeon backend says 58%~ (edit: now 87% Some engines even say they get 97.5 % hardware errors)
3) my cgminer is already fully updated since the latest version came with Minepeon 0.2.3

cgminer has been updated a few times since 0.2.3 came out .

if you want the latest try the instructions for updating the software via a git pull above.

I think there was a BFL fix of some sort in the 3.4.2 version.

[Grouver (BtcBalance)]

Well its weird. When I screen -r I see bfgminer and cgminer my hardware errors are just around 2%~
But when I check out Minepeon it says the average of both singles is 51%.
With bfgminer some engines even display 98% hardware errors. Not sure whats going on.





Edit:

via a git pull above.

Isnt this a bit dangerous and only for advanced users?
I don't want to corrupt my entire Minepeon system.
Or is this bug fixed now that it crashed when git pull failed.

[KyrosKrane]

If you get an error when doing git pull, try this.  It should fix it.

Log in, then run these commands:

Code:

sudo -i
find /opt/minepeon -user root | xargs chmod a+w
exit

Type your password when prompted by sudo. Ignore the error message that gets spit out from the find command. Then do git pull as normal, and it works.

[segabtc]

having same problem as above, cgminer reports 1 to 2 % HW errors but Peon status page report 40 % HW errors. 0.2.3 with cgminer 3.4.2. could this just be a reporting error? Have 2 bfl 60's hooked up to minepeon and run great. Checked with pool and it doesnt appear I am losing shares.... Dont know what to think about this.

Anyway great job of minepeon!!!! this is a plug and play system!

[LogicalUnit]

Well its weird. When I screen -r I see bfgminer and cgminer my hardware errors are just around 2%~
But when I check out Minepeon it says the average of both singles is 51%.
With bfgminer some engines even display 98% hardware errors. Not sure whats going on.

I'm getting the same issue, and I think there is a bug in the number of accepted shares being reported. Reported shares on the web UI are roughly one tenth of actual shares. Maybe someone's deleting a zero somewhere. . .

[tk1337]

Well its weird. When I screen -r I see bfgminer and cgminer my hardware errors are just around 2%~
But when I check out Minepeon it says the average of both singles is 51%.
With bfgminer some engines even display 98% hardware errors. Not sure whats going on.

I'm getting the same issue, and I think there is a bug in the number of accepted shares being reported. Reported shares on the web UI are roughly one tenth of actual shares. Maybe someone's deleting a zero somewhere. . .

It does appear that my accepted shares are off by about 1300 / per miner, from cgMiner's stats to the stats being displayed on MinePeon's webUI.
Since I'm only running erupters, I could see where something on a larger scale could report much higher.

Going to take a look at the code since I'm awake (and shouldn't be).

[tk1337]

Well its weird. When I screen -r I see bfgminer and cgminer my hardware errors are just around 2%~
But when I check out Minepeon it says the average of both singles is 51%.
With bfgminer some engines even display 98% hardware errors. Not sure whats going on.

I'm getting the same issue, and I think there is a bug in the number of accepted shares being reported. Reported shares on the web UI are roughly one tenth of actual shares. Maybe someone's deleting a zero somewhere. . .

It does appear that my accepted shares are off by about 1300 / per miner, from cgMiner's stats to the stats being displayed on MinePeon's webUI.
Since I'm only running erupters, I could see where something on a larger scale could report much higher.

Going to take a look at the code since I'm awake (and shouldn't be).

Okay, I wrote a separate connector to the cgMiner API (just to rule out anything currently in MinePeon)... and it appears that the response from CGMiner is actually returning different data than what it's showing on the output screen. Could be a bug in the latest CGMiner release.

Edit: With 623 pages on the CGMiner thread, not sure if the issue has already been reported, but definitely seeing a 1319 accepted share difference for each USB Erupter I have going at the moment. However I made a post there, we'll see if anything becomes of it, it could be some kind of offset (maybe?) or just a bug in the API callback.

[LogicalUnit]

Okay, I wrote a separate connector to the cgMiner API (just to rule out anything currently in MinePeon)... and it appears that the response from CGMiner is actually returning different data than what it's showing on the output screen. Could be a bug in the latest CGMiner release.

Edit: With 623 pages on the CGMiner thread, not sure if the issue has already been reported, but definitely seeing a 1319 accepted share difference for each USB Erupter I have going at the moment. However I made a post there, we'll see if anything becomes of it, it could be some kind of offset (maybe?) or just a bug in the API callback.

If it's any help, my LCDStats screen is outputting the correct figures.

[tk1337]

Okay, I wrote a separate connector to the cgMiner API (just to rule out anything currently in MinePeon)... and it appears that the response from CGMiner is actually returning different data than what it's showing on the output screen. Could be a bug in the latest CGMiner release.

Edit: With 623 pages on the CGMiner thread, not sure if the issue has already been reported, but definitely seeing a 1319 accepted share difference for each USB Erupter I have going at the moment. However I made a post there, we'll see if anything becomes of it, it could be some kind of offset (maybe?) or just a bug in the API callback.

If it's any help, my LCDStats screen is outputting the correct figures.

hm... that is interesting.

I'll work on this more tomorrow, I need some sleep.

[tk1337]

okay, need to go to sleep, I keep screwing with this...

I changed the parameter to the API call to a integer (1) rather than a boolean (true) and now I'm getting a different set of data back that is actually correct.

Gives me something to go off of, I'll look at it more in the morning.

[kano]

okay, need to go to sleep, I keep screwing with this...

I changed the parameter to the API call to a integer (1) rather than a boolean (true) and now I'm getting a different set of data back that is actually correct.

Gives me something to go off of, I'll look at it more in the morning.

I've got no idea what you are referring to there.
I wrote the cgminer API, but yeah I need a little more information to make any sense of that.

[tk1337]

okay, need to go to sleep, I keep screwing with this...

I changed the parameter to the API call to a integer (1) rather than a boolean (true) and now I'm getting a different set of data back that is actually correct.

Gives me something to go off of, I'll look at it more in the morning.

I've got no idea what you are referring to there.
I wrote the cgminer API, but yeah I need a little more information to make any sense of that.

Basically, when passing the command & parameter to the API from PHP, it seems to like an integer to be passed rather than a boolean. Something PHP might be doing odd on the write to the socket, who knows, some times there is some strange shit that pops up.

Code:

$command = 'summary';
$parameter= 1;

Rather than

Code:

$command = 'summary';
$parameter= true;

PHP will parse 1 as true (and 0 as false) unless explicitly defined...
ie:

if($foo == 1)() is the same as if($foo == true){} (also if($foo == '1'){} and if($foo == "1")() even though technically the 1 is a string in the last example, PHP doesn't really care), unless you specifically look for a boolean like so, if($foo === true){} or if(is_bool($foo) && $foo == true){}

When I'm passing it to the API socket as true, it seems to not like it for some reason, but it's fine with it as an integer, someone on the CGMiner thread did bring up an interesting thing about SDCard read/writes and I have about 10 SDCards at the house, so I may clone the card and just check that. I'm at work now, if I survive the day with no sleep, I'll try a few things and post more info later today.

[kano]

Ah OK some issue in some PHP you wrote - nothing to do with the API.
(the entire API read/write is text)

miner.php (which I wrote) doesn't have such problems.

Also as I stated before regarding SD cards:
https://bitcointalk.org/index.php?topic=137934.msg2673249;topicseen#msg2673249