Avalon unit keeps restarting any help ?

[thorvald]

the problem was not fixed
the hub was replaced
the psu was replaced
it restarts random

When the unit stops there is this in kernal logs this

Mar 20 19:24:38 OpenWrt2 user.info sysinit: 1156
Mar 20 19:24:42 OpenWrt2 kern.err kernel: [   41.870000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:43 OpenWrt2 kern.err kernel: [   43.580000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:45 OpenWrt2 kern.err kernel: [   45.280000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:47 OpenWrt2 kern.err kernel: [   46.980000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:50 OpenWrt2 kern.err kernel: [   50.390000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:52 OpenWrt2 kern.err kernel: [   52.090000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:55 OpenWrt2 kern.err kernel: [   55.510000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:57 OpenWrt2 kern.err kernel: [   57.210000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:24:59 OpenWrt2 kern.err kernel: [   58.910000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:00 OpenWrt2 kern.err kernel: [   60.620000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:04 OpenWrt2 kern.err kernel: [   64.000000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:06 OpenWrt2 kern.err kernel: [   65.700000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:07 OpenWrt2 kern.err kernel: [   67.400000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:09 OpenWrt2 kern.err kernel: [   69.110000] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006400
Mar 20 19:25:09 OpenWrt2 kern.err kernel: [   69.110000] ath: phy0: Could not stop RX, we could be confusing the DMA engine when we start RX up
Mar 20 19:25:09 OpenWrt2 kern.err kernel: [   69.130000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:10 OpenWrt2 cron.err crond[1088]: time disparity of 64 minutes detected
Mar 20 19:25:11 OpenWrt2 kern.err kernel: [   70.840000] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006400
Mar 20 19:25:11 OpenWrt2 kern.err kernel: [   70.840000] ath: phy0: Could not stop RX, we could be confusing the DMA engine when we start RX up
Mar 20 19:25:11 OpenWrt2 kern.err kernel: [   70.860000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:14 OpenWrt2 kern.err kernel: [   74.250000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:16 OpenWrt2 kern.err kernel: [   75.950000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:21 OpenWrt2 kern.err kernel: [   81.010000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:24 OpenWrt2 kern.err kernel: [   84.390000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:26 OpenWrt2 kern.err kernel: [   86.110000] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006400
Mar 20 19:25:26 OpenWrt2 kern.err kernel: [   86.110000] ath: phy0: Could not stop RX, we could be confusing the DMA engine when we start RX up
Mar 20 19:25:26 OpenWrt2 kern.err kernel: [   86.130000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:29 OpenWrt2 kern.err kernel: [   89.510000] ath: phy0: Failed to stop TX DMA, queues=0x001!
Mar 20 19:25:31 OpenWrt2 kern.err kernel: [   91.230000] ath: phy0: DMA failed to stop in 10 ms AR_CR=0x00000024 AR_DIAG_SW=0x02000020 DMADBG_7=0x00006400
Mar 20 19:25:31 OpenWrt2 kern.err kernel: [   91.230000] ath: phy0: Could not stop RX, we could be confusing the DMA engine when we start RX up
Mar 20 19:25:31 OpenWrt2 kern.err kernel: [   91.250000] ath: phy0: Failed to stop TX DMA, queues=0x001!

[1714199154]

1714199154

[1714199154]

1714199154

[1714199154]

1714199154

[needbmw]

Does it connected to network via wi-fi or ethernet port?

[kaerf]

at one point i got those ath: phy0 errors too. after a second reflash of firmware without saving configuration those errors went away.

the thing that catches my eye in the logs is the crond error. make sure your avalon has outbound network access on UDP port 123. also delete /tmp/cm.log for good measure.

[Gator-hex]

Updated firmware?
https://bitcointalk.org/index.php?topic=148212.60

[eleuthria]

Are you mining on a stratum pool that supports vardiff or minimum difficulty settings?  When jgarzik was first testing the Avalon, BTC Guild was the most stable/reliable pool with regards to unit restarts due to an aggressive variable difficulty.  Since then many other pools have updated to include variable difficulty and/or manual settings.  An Avalon should be running at 32+ difficulty.  Higher produced even more stability in the unit, but obviously more variance per 24 hour period.

I don't believe this is still a problem with the new firmware (posted above), but it's something to think about.  Make sure you're on a pool that will let you adjust to higher difficulties, or has vardiff.

[thorvald]

i ll try it your way kaerf
the protocol is stratum difficulty min 32 on btcguild
no overclock on the system
firmware is the version 225
i`ve tested also the latest version 20 but the cgminer did not start
i do not have access to the system now,it is in the data center i`ll do it tomorrow morning

[Nexus2k]

...
i do not have access to the system now,it is in the data center i`ll do it tomorrow morning

Does the datacenter has its own firewalls which could prevent outgoing network traffic to the pools ?

[vdragon]

Check the firewalls, that solved my problems

[johnyj]

Have you first tried to disable that cgminer-monitor script or increase the check interval and see what will happen?

I also had at least one restart per day, I reduced the check frequency from 5 to 15, and it is stable since then

System->scheduled tasks
*/15 *  *   *   *     /usr/bin/cgminer-monitor

[thorvald]

if you can please tell me how i`ll do it
10x

[loshia]

Have you first tried to disable that cgminer-monitor script or increase the check interval and see what will happen?

I also had at least one restart per day, I reduced the check frequency from 5 to 15, and it is stable since then

What about your HW error rate you are having?

I will post some interesting rrd graphs tomorrow. It might turn out that a unit can recover alone but it needs a time. The question still remains why is this happening at all. Anyway you will see the graphs tomorrow. It is too late now


I think GBT can cause the restarts also (out of memory). My unit had only 3M free ram with all stuff running and stratum. I know that GBT uses a lot more of ram in general.
 

[cypherdoc]

Have you first tried to disable that cgminer-monitor script or increase the check interval and see what will happen?

I also had at least one restart per day, I reduced the check frequency from 5 to 15, and it is stable since then

What about your HW error rate you are having?

I will post some interesting rrd graphs tomorrow. It might turn out that a unit can recover alone but it needs a time. The question still remains why is this happening at all. Anyway you will see the graphs tomorrow. It is too late now


I think GBT can cause the restarts also (out of memory). My unit had only 3M free ram with all stuff running and stratum. I know that GBT uses a lot more of ram in general.
 

didn't someone say there was a way to increase that RAM?

[thorvald]

after more then 30 restarts in the day

the unit didn`t restart any more
i`ve done nothing to it
+4 btc after last night 
i can`t explain it

[xiangfu]

Should be fixed in latest firmware(20130321)
  https://en.bitcoin.it/wiki/Avalon#20130321

[thorvald]

i`ll load and test it after the first reset (still going on 12h+)
in the version 320 the cgminer did not start so i reverted to 255
it worked just in debug mod the cgminer started with ~0.9.gh 

10x
Thorvald

[invader]

Thanks for update! Testing new firmware.
One issue i noticed - when i tried to set API Allow "W:0.0.0.0/0", internal & external monitoring isn't working.
So i changed it to "W:127.0.0.1,192.168.0.0/16" to allow my own monitoring scripts.
Second minor issue - with dnsmasq option "domain-needed" my internal network plain hostnames (without domain) isn't working.

[thorvald]

after 22h+ the first restart i`ve updated the firmware to the latest version

after roi donation will be made
10x
thorvald