Erupter Blade not submitting work at the expected rate

AustenN

newbie

Activity: 26

Merit: 0

Why is your proxy saying you are accepting shares at difficulty 1? I just set up my blade and my proxy defaults to difficulty 9. I am getting 11 Ghash ATM.

Theo1355

newbie

Activity: 46

Merit: 0

Guys, for those of you who like fighting with the command line shell, use stratum. For everyone else, just use getwork or gbt protocals.

pidge

member

Activity: 101

Merit: 10

Hmm, I may have solved it by specifying a custom user and password for share submission:

Code:

mining_proxy.exe -o stratum.btcguild.com -p 3333 -gp 8342 -sp 3342 -cu pjnield_4 -cp 1234

how bloody annoying

pidge

member

Activity: 101

Merit: 10

Pointed a blade at p2pool, and at least it did some work, and reached ~3GH. So, not the blade? something funky with my mining_proxy version or command line?

Command line (I've used various parameters)

Code:

mining_proxy.exe -o stratum.btcguild.com -p 3333 -gp 8332 -rt

Version (latest as far as I know)

Code:

2013-09-18 00:45:16,124 WARNING proxy mining_proxy.main # Stratum proxy version: 1.5.2

pidge

member

Activity: 101

Merit: 10

That response is 732 bytes. I wonder if a few more bytes could be stritpped...

pidge

member

Activity: 101

Merit: 10

Hmmm, a comment in the code at https://github.com/forrestv/p2pool/blob/master/p2pool/bitcoin/worker_interface.py

Code:

# ASICMINER BE Blades apparently have a buffer overflow bug and
# can't handle much extra in the getwork response

Here is a response I captured from mining_proxy.exe (without midstate, shorted the x-stratum response header)

Code:

HTTP/1.1 200 OK
Transfer-Encoding: chunked
X-Roll-Ntime: 1
X-Long-Polling: /lp
Server: TwistedWeb/12.2.0
X-Stratum: 123
Date: Tue, 17 Sep 2013 12:07:44 GMT
Content-Type: application/json
20b
{"error": null, "id": 1, "result": {"hash1": "00000000000000000000000000000000000000000000000000000000000000000000008000000000000000000000000000000000000000000000000000010000", "data": "0000000246dc88cdb3254ff6fa084275593d40a5fc8570812fc9e7bf0000000c00000000ef69fc52bcde11419e278f31dea391be6682a632ca44af9269b792eeeffab2755238460b1926222200000000000000800000000000000000000000000000000000000000000000000000000000000000000000000000000080020000", "target": "0000000000000000000000000000000000000000000000000000ffff00000000"}}
0

(edit: oops, wrong response!)

xsfgsdrwe

newbie

Activity: 56

Merit: 0

Quote from: pidge on September 15, 2013, 09:30:38 AM

I have two BE Blades running, but at around 2013-09-16 08:30 UTC they stopped regularly completing work - the would ask for work, but not submitt many completed shares

I've done the following, with no change to the behaviour:
- Restarting the blades (from web page and power cycle)
- Restarted mining_proxy
- Factory reset of a blade and reconfigured
- Ran with just one blade (to try to eliminate power issues)
- Swapped in another new blade.

Voltage from the PSU is still good (12.4V), but I've got no means of checking for excess noise on the power. The PSU is a re-purposed HP Proliant G4 hot-swap PSU (12V, 47A).

I'm down to either the PSU has gone faulty, or there's something odd with the work being supplied. I might have a look at the output on the debug port, which will be interesting as I've only got a logic analyzer, not a USB to serial TTL 3.3V to hand.

Any ideas?

Observations:
- The "chips" status is all O's, no X's
- The blades request 8 shares to work on at a time, and may submit a completed share every now and then.
- Example mining_proxy output below, normally the screen just scrolls and scrolls past.

Code:

2013-09-16 00:20:31,878 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:31,953 INFO proxy jobs.submit # Submitting 25251d53
2013-09-16 00:20:31,969 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,019 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,036 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,112 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,135 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,174 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,190 WARNING proxy getwork_listener._on_submit # [230ms] Share from '' accepted, diff 1
2013-09-16 00:20:32,288 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,303 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,319 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,394 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,418 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,437 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,513 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,536 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,624 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,755 INFO proxy jobs.submit # Submitting f226d12f
2013-09-16 00:20:32,792 INFO proxy jobs.submit # Submitting def7eb60
2013-09-16 00:20:32,806 INFO proxy jobs.submit # Submitting 837ac763
2013-09-16 00:20:32,816 INFO proxy jobs.submit # Submitting 9db12769
2013-09-16 00:20:32,990 WARNING proxy getwork_listener._on_submit # [227ms] Share from '' accepted, diff 1
2013-09-16 00:20:33,414 WARNING proxy getwork_listener._on_submit # [615ms] Share from '' accepted, diff 1
2013-09-16 00:20:33,421 WARNING proxy getwork_listener._on_submit # [606ms] Share from '' accepted, diff 1
2013-09-16 00:20:33,433 WARNING proxy getwork_listener._on_submit # [609ms] Share from '' accepted, diff 1
2013-09-16 00:20:48,844 INFO proxy client_service.handle_event # New job 22520 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:21:18,857 INFO proxy client_service.handle_event # New job 22528 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:21:48,489 INFO proxy getwork_listener._on_authorized # Worker 'pjnield_5' asks for new work
2013-09-16 00:21:48,558 INFO proxy getwork_listener._on_authorized # Worker 'pjnield_5' asks for new work
2013-09-16 00:21:49,503 INFO proxy client_service.handle_event # New job 22532 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:22:18,858 INFO proxy client_service.handle_event # New job 22540 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:22:48,861 INFO proxy client_service.handle_event # New job 22556 for prevhash b9303f27, clean_jobs=False

Seems the problem may be in the devices as btcguild is usually on par with things like this.

They have a warranty with free shipping if the devices is in fact flawed.

pidge

member

Activity: 101

Merit: 10

Probably the same problem as here https://bitcointalk.org/index.php?topic=293787.new, and as reported in by a number of users starting here https://bitcointalk.org/index.php?topic=205369.580

pidge

member

Activity: 101

Merit: 10

I use BTC-Guild. I've also tried Slush's pool, same symptoms.

HellDiverUK

hero member

Activity: 1246

Merit: 501

What pool are you connecting to?

pidge

member

Activity: 101

Merit: 10

I have two BE Blades running, but at around 2013-09-16 08:30 UTC they stopped regularly completing work - the would ask for work, but not submitt many completed shares

I've done the following, with no change to the behaviour:
- Restarting the blades (from web page and power cycle)
- Restarted mining_proxy
- Factory reset of a blade and reconfigured
- Ran with just one blade (to try to eliminate power issues)
- Swapped in another new blade.

Voltage from the PSU is still good (12.4V), but I've got no means of checking for excess noise on the power. The PSU is a re-purposed HP Proliant G4 hot-swap PSU (12V, 47A).

I'm down to either the PSU has gone faulty, or there's something odd with the work being supplied. I might have a look at the output on the debug port, which will be interesting as I've only got a logic analyzer, not a USB to serial TTL 3.3V to hand.

Any ideas?

Observations:
- The "chips" status is all O's, no X's
- The blades request 8 shares to work on at a time, and may submit a completed share every now and then.
- Example mining_proxy output below, normally the screen just scrolls and scrolls past.

Code:

2013-09-16 00:20:31,878 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:31,953 INFO proxy jobs.submit # Submitting 25251d53
2013-09-16 00:20:31,969 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,019 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,036 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,112 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,135 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,174 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,190 WARNING proxy getwork_listener._on_submit # [230ms] Share from '' accepted, diff 1
2013-09-16 00:20:32,288 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,303 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,319 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,394 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,418 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,437 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,513 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,536 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,624 INFO proxy getwork_listener._on_authorized # Worker '' asks for new work
2013-09-16 00:20:32,755 INFO proxy jobs.submit # Submitting f226d12f
2013-09-16 00:20:32,792 INFO proxy jobs.submit # Submitting def7eb60
2013-09-16 00:20:32,806 INFO proxy jobs.submit # Submitting 837ac763
2013-09-16 00:20:32,816 INFO proxy jobs.submit # Submitting 9db12769
2013-09-16 00:20:32,990 WARNING proxy getwork_listener._on_submit # [227ms] Share from '' accepted, diff 1
2013-09-16 00:20:33,414 WARNING proxy getwork_listener._on_submit # [615ms] Share from '' accepted, diff 1
2013-09-16 00:20:33,421 WARNING proxy getwork_listener._on_submit # [606ms] Share from '' accepted, diff 1
2013-09-16 00:20:33,433 WARNING proxy getwork_listener._on_submit # [609ms] Share from '' accepted, diff 1
2013-09-16 00:20:48,844 INFO proxy client_service.handle_event # New job 22520 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:21:18,857 INFO proxy client_service.handle_event # New job 22528 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:21:48,489 INFO proxy getwork_listener._on_authorized # Worker 'pjnield_5' asks for new work
2013-09-16 00:21:48,558 INFO proxy getwork_listener._on_authorized # Worker 'pjnield_5' asks for new work
2013-09-16 00:21:49,503 INFO proxy client_service.handle_event # New job 22532 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:22:18,858 INFO proxy client_service.handle_event # New job 22540 for prevhash b9303f27, clean_jobs=False
2013-09-16 00:22:48,861 INFO proxy client_service.handle_event # New job 22556 for prevhash b9303f27, clean_jobs=False

Topic: Erupter Blade not submitting work at the expected rate (Read 2002 times)