WATCHDOG: GPU 3 hangs in OpenCL call, exit

[marcopollo68]

hello, since yesterday I change windows for HiveOS and I’m very happy with it but I get a few errors with my GPU 3. I get 1 or 2 crashes in an hour and always the same error message: Claymore Reboot: WATCHDOG: GPU 3 hangs in OpenCL call, exit
I have tried to change riser nothing change, I have set 0 Overclock the same issue. I have tried to flash an other BIOS and always the same, sometimes it keep mining longer than before, but always after 1-2 hour it will crash again beacause of my GPU 3. I let your le message error here:
=== Last 50 lines of /hive/claymore106/lastrun_reboot.log ===

04:58:16:830 5cff9700 parse packet: 36
04:58:16:830 5cff9700 DCR: Share accepted (245 ms)!
04:58:16:830 5cff9700 new buf size: 0
04:58:17:128 5cff9700 got 402 bytes
04:58:17:128 5cff9700 buf: {“id”:null,“method”:“mining.notify”,“params”:[“b36f”,“07a6c10fa1ea7f96a927d48f9f3a0d7eca18a5679deda9550000002300000000”,“d0fe3e6185253ab86e9bca246b9fbb97e8064a029821ac6b8dd0cd0bbcc9cc65884722e734dd93749c851266965756826ef208293a70aade98ef95bb30edeae201004d99aa1ca16305000c004fa000006e1545191b8fefd101000000b72c0300dd270000d89b6e5a0000000000000000”,“05000000”,[],“05000000”,“1945156e”,“5a6e9bd8”,false]}

04:58:17:128 5cff9700 parse packet: 401
04:58:17:129 5cff9700 new buf size: 0
04:58:17:129 5cff9700 DCR: 01/29/18-04:58:17 - New job from dcr.suprnova.cc:3252
04:58:17:129 5cff9700 target: 0x000000001b6b2f72 (diff: 40GH), block #208055
04:58:17:518 667fc700 DCR: put share nonce 2ee95bd enonce 4c6a000f
04:58:17:518 667fc700 DCR round found 1 shares
04:58:17:519 5cff9700 DCR: 01/29/18-04:58:17 - SHARE FOUND - (GPU 1)
04:58:17:519 5cff9700 send: {“params”: [“marcopollo68.marcminer01”, “b36f”, “94a885624c6a000f75324c6f”, “5a6e9bd8”, “bd95ee02”], “id”: 11, “method”: “mining.submit”}

04:58:18:291 5cff9700 got 37 bytes
04:58:18:291 5cff9700 buf: {“id”:11,“result”:true,“error”:null}

04:58:18:291 5cff9700 parse packet: 36
04:58:18:291 5cff9700 DCR: Share accepted (772 ms)!
04:58:18:291 5cff9700 new buf size: 0
04:58:18:724 88c92700 GPU 0 temp = 67, old fan speed = 22, new fan speed = 25

04:58:18:724 88c92700 GPU 1 temp = 64, old fan speed = 22, new fan speed = 25

04:58:18:724 88c92700 GPU 2 temp = 63, old fan speed = 22, new fan speed = 25

04:58:18:724 88c92700 GPU 3 temp = 56, old fan speed = 22, new fan speed = 25

04:58:18:725 88c92700 GPU 4 temp = 60, old fan speed = 22, new fan speed = 25

04:58:18:725 88c92700 GPU 5 temp = 62, old fan speed = 22, new fan speed = 25

04:58:18:725 88c92700 GPU0 t=67C fan=22%%, GPU1 t=64C fan=22%%, GPU2 t=63C fan=22%%, GPU3 t=56C fan=22%%, GPU4 t=60C fan=22%%, GPU5 t=62C fan=22%%
04:58:18:734 37fff700 srv_thr cnt: 1, IP: 127.0.0.1
04:58:18:734 37fff700 recv: 51
04:58:18:734 37fff700 srv pck: 50
04:58:18:734 37fff700 srv bs: 0
04:58:18:734 37fff700 sent: 263
04:58:20:736 89493700 em hbt: 0, dm hbt: 0, fm hbt: 91,
04:58:20:736 89493700 watchdog - thread 0 (gpu0), hb time 140
04:58:20:736 89493700 watchdog - thread 1 (gpu0), hb time 69
04:58:20:736 89493700 watchdog - thread 2 (gpu1), hb time 69
04:58:20:736 89493700 watchdog - thread 3 (gpu1), hb time 143
04:58:20:736 89493700 watchdog - thread 4 (gpu2), hb time 1
04:58:20:736 89493700 watchdog - thread 5 (gpu2), hb time 73
04:58:20:736 89493700 watchdog - thread 6 (gpu3), hb time 86871
04:58:20:736 89493700 WATCHDOG: GPU 3 hangs in OpenCL call, exit
04:58:20:736 89493700 watchdog - thread 7 (gpu3), hb time 86946
04:58:20:736 89493700 WATCHDOG: GPU 3 hangs in OpenCL call, exit
04:58:20:736 89493700 watchdog - thread 8 (gpu4), hb time 69
04:58:20:736 89493700 watchdog - thread 9 (gpu4), hb time 139
04:58:20:736 89493700 watchdog - thread 10 (gpu5), hb time 46
04:58:20:736 89493700 watchdog - thread 11 (gpu5), hb time 116
04:58:20:736 89493700 Rebooting
I hope some one can help me thanks.

[Meateater]

Hi!
Did you flash the card using HiveOS or did you use Windows/ ATI Flasher? I have noticed some strange things when I use the flashing tool in HiveOS.
Usually “hangs in OpenCL call” comes when you overclocked too much. Perhaps you should try to flash in Windows if you haven’t already.

[Stef]

Hi

Yes Windows ethos or hive os, this message is 99% of the time because too much memory clock OR bios. The exact same card will do it if i put more then 1825 in memory . BIG difference, my Windows was freezing where hive os just restart a new miner and voila it continues to mine !!!

cheers

[marcopollo68]

So I have some news, I tried to reflash the card and nothing change. I have try without overclocking and I get always the same problem. I have decied to get back on simple eth mining and with simple mining it works fine! So I found this interessing and I have tried dual mining on my other rig with gtx 1070 and I get exactly the same issue, there is always one GPU which mine fine at the beginning and after 10-15 min it drops down to zero and rig reboot. So I think there is a really problem with dual mining on HiveOS because before that I use HiveOS it was mining fine on windows, and I have tried SIA and DECRED and the two get this problem!
Can some other people who make dual mining on HiveOS share their experience please. Did you get the same issue or it works fine for you?

[Zep]

I’ve tried dual mining, but temperatures were too high. I had card errors and reboot too, but not after only 10minutes, I was able to dual mine for almost a day without errors. I don’t think it’s something wrong with hive, I think that if you want to dual mine you have to find the right settings for you cards, maybe you are overclocking too much.

[Tj_Pineda]

i think i have resolve the watchdog problem. mine also reboots, so
i go to the shellinthebox type some commands. then its ok

[wkholi]

did you get an error like this?

How did you resolve it?