Problem: Avalon#1 stop working very often

[Bitbird]

I'm trying to work on the abnormal problem of the Avalon#1 which cause it stop mining very often (around every few hours or more).

The Avalon was connected to Linux PC and mining with cgminer 2.10.5 avalon branch. The message "AVA0: invalid nonce -HW error" appears frequently when mining (see: https://i.imgur.com/7d2SM9H.jpg ). I don't know whether it's related to the problem or not.

When the cgminer stop working and showing the continuous error messages (I'll copy them when it happen again.) Restart cgminer can solved the problem. But if this procedure can't be automated, large loss would be expected.

Anyone encountered similar problem? Is there any solution? Thanks!

p.s. I use latest firmware: openwrt-ar71xx-generic-tl-wr703n-v1-squashfs-factory.bin   27-Feb-2013 20:09    3.8M

[Bogart]

I'm mining using the Avalon's internal host and don't have a problem with it hanging, but I do notice that cgminer seems to restart itself fairly often.  Elapsed Time shown on the cgminer status page is usually quite low.

I think they may have some kind of a watchdog that restarts it.

[Gator-hex]

Did you try reading this thread?

[FIXED] Avalon URGENT ISSUE: both of my avalon do not work any more
https://bitcointalk.org/index.php?topic=148212.0

[Bitbird]

@Bogart

Since I can't got the wr703n working I may need a watchdog or daemon on the PC side. When the daemon detected the error it can restart cgminer. But I don't know how to build one for now.


@Gator-hex

Not yet! Will read it. Thanks!

[techwtf]

0.5-0.75% HW Error is normal. as one avalon could do 900+ nonces in a minute, ~6 HW Errors per minute is quite normal. you should check "devs" api for "MHS 5s" or "Diff1 Work" (should be increasing). there is a similar monitoring script inside the default firmware using "Accepted" as metric (using "stats" api), see https://github.com/BitSyncom/cgminer-openwrt-packages/blob/master/cgminer/files/cgminer-monitor .

# for RESTART_MONITOR, see https://bitcointalk.org/index.php?topic=145885.0
# but I do NOT guarantee it could handle every issue.
IP=xxx.xxx.xxx.xxx
PORT=4028
RESTART_MONITOR="(echo -n 'devs' | nc ${IP} ${PORT} | grep 'MHS 5s=0.0') && (echo -n 'restart' | nc ${IP} ${PORT})"
echo $RESTART_MONITOR
while true; do date; sleep 10; sh -c "$RESTART_MONITOR"; done

[Bitbird]

@techwtf

I see.  Although not completely understand the later part.
Will doing some research. Thanks!

[Bogart]

Can't make the WR703N work huh?  Is it totally unresponsive?  Does the light come on?

You can get a new WR703N for ~$20.  I have several sitting beside me here.

[loshia]

Mine us up for 24 hours without any issues. However i belive that Avalon usb is very sensative to radiation from psu. can you mount some metal box above usb chip (controler board) to act as shield or try to move controler as far as you can from psu?

it might help

[ProfMac]

Can't make the WR703N work huh?  Is it totally unresponsive?  Does the light come on?

You can get a new WR703N for ~$20.  I have several sitting beside me here.

How would it make financial sense for me to order a handful of the TP-LINK TL-WR703N v1 and flash them for backups before the community needs them?

[Bogart]

Can't make the WR703N work huh?  Is it totally unresponsive?  Does the light come on?

You can get a new WR703N for ~$20.  I have several sitting beside me here.

How would it make financial sense for me to order a handful of the TP-LINK TL-WR703N v1 and flash them for backups before the community needs them?

I suppose that depends on the expected rate of failure, and what kind of premium you think someone would pay for a turnkey replacement.

I have mine around just to play with.

[ProfMac]

Can't make the WR703N work huh?  Is it totally unresponsive?  Does the light come on?

You can get a new WR703N for ~$20.  I have several sitting beside me here.

How would it make financial sense for me to order a handful of the TP-LINK TL-WR703N v1 and flash them for backups before the community needs them?

I suppose that depends on the expected rate of failure, and what kind of premium you think someone would pay for a turnkey replacement.

I have mine around just to play with.

I ordered one from Amazon for about $20.  It should be here by Wed. 

[libertybuck]

I ordered one from Amazon for about $20.  It should be here by Wed. 

Sir, the WR703N shipped with Avalon should be a little different with yours. It is modified with larger RAM and Flash. And a resistor is mounted at different position. 

[libertybuck]

One thing want to share with friends is that, avalon is a professional mining machine, to avoid destory it accidently please do not PLAY with avalon. If not necessary please do not open it's case. If you opened it you should not power on it for a long time. Leave case open could only make components over-heat.

[ProfMac]

I ordered one from Amazon for about $20.  It should be here by Wed. 

Sir, the WR703N shipped with Avalon should be a little different with yours. It is modified with larger RAM and Flash. And a resistor is mounted at different position. 

Thank you.
Do you know whether the modifications done at manufacturing time, or in the field?

[libertybuck]

I ordered one from Amazon for about $20.  It should be here by Wed. 

Sir, the WR703N shipped with Avalon should be a little different with yours. It is modified with larger RAM and Flash. And a resistor is mounted at different position. 

Thank you.
Do you know whether the modifications done at manufacturing time, or in the field?

check here please:

the 703N is a modded ver, we moved a 0R resistor from one place to another to let the 703N could get power form its USB-A port.

[Bogart]

There are videos all over youtube of ppl doing these mods.

[Bogart]

I'm mining using the Avalon's internal host and don't have a problem with it hanging, but I do notice that cgminer seems to restart itself fairly often.  Elapsed Time shown on the cgminer status page is usually quite low.

I think they may have some kind of a watchdog that restarts it.

I'm now finding mining using stratum to be far more stable.  4h18m elapsed is showing now on the cgminer status page.

[Bitbird]

@loshia

True! Avalons' USB is quite sensitive. Sometime when I switch the light (I use LED light) of the room it will stop function and need re plug-in for restart. I already putting PSU (changed to a better one) outside the case but the situation remain the same.

[JerimiahJohnson]

Thank you for working on this, please if you or anyone who is good with Avalons / may have a fix read my post here:

https://bitcointalk.org/index.php?topic=156323.msg1696891#msg1696891

Its putting itself into idle mode for some reason,will gladly pay bounty for fix

thank you.

-JJ