claymore miner: OpenCL error -61 - cannot allocate big buffer for DAG.

[sulla-x]

Dear all!

I need some help with the claymore miner:
I have a setup with 4 MSI Radeon RX580 Armor 8G OC cards in Linux Ubuntu 16.04.02 using AMDGPU-PRO 17.10 drivers and claymore 9.3 software, 8GB Ram + 8GB swap. 2 cards are directly on the MB (x16, x4 PCIe) and 2 with risers (x1 ==> x16).

The system starts fine and mines for 16-20 hours without problems, all cards run fine.
Suddenly, without any errors on the console or the logfile, 3 cards drop out and only one card continues mining.

Then, claymore miner restarts automatically, now all 4 cards are disfunct and it delivers
OpenCL error -61 - cannot allocate big buffer for DAG
because for some reason, it doesn't recognise the full 8GB of the graphics cards memory any more, but only < 2GB. Strange.

After a reboot everything comes up normally for another 12-20 hrs.

Can somebody point me to a possible solution?

Thanx for support!


here's console output from first startup until mining starts, then at the time mining stops and the miner restarts again:


============== console output ==================

▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ͻ
▒      Claymore's Dual ETH + DCR/SC/LBC/PASC GPU Miner v9.3      ▒
▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ͼ

ETH: 3 pools are specified
Main Ethereum pool is eu2.ethermine.org:4444
DCR: 2 pools are specified
Main Decred pool is yiimp.ccminer.org:3252

Catalyst 15.12 is REQUIRED for best performance and compatibility
For Polaris cards, use latest drivers
At least 16 GB of Virtual Memory is required for multi-GPU systems
Make sure you defined GPU_MAX_ALLOC_PERCENT 100
Be careful with overclocking, use default clocks for first tests
Press "s" for current statistics, "0".."9" to turn on/off cards, "r" to reload pools, "e" or "d" to select current pool
OpenCL initializing...

AMD Cards available: 4
GPU #0: Ellesmere, 8167 MB available, 36 compute units
GPU #0 recognized as Radeon RX 480/580
GPU #1: Ellesmere, 8169 MB available, 36 compute units
GPU #1 recognized as Radeon RX 480/580
GPU #2: Ellesmere, 8169 MB available, 36 compute units
GPU #2 recognized as Radeon RX 480/580
GPU #3: Ellesmere, 8169 MB available, 36 compute units
GPU #3 recognized as Radeon RX 480/580
POOL/SOLO version
GPU #0: algorithm ASM
GPU #1: algorithm ASM
GPU #2: algorithm ASM
GPU #3: algorithm ASM
No NVIDIA CUDA GPUs detected.
Total cards: 4
AMD ADL library not found.
ETH: Stratum - connecting to 'eu2.ethermine.org' <91.121.222.33> port 4444
ETH: Stratum - Connected (eu2.ethermine.org:4444)
DUAL MINING MODE ENABLED: ETHEREUM+DECRED
ETH: qtminer stratum mode
Watchdog enabled
Remote management (READ-ONLY MODE) is enabled on port 3333

ETH: Authorized
 DCR: Stratum - connecting to 'yiimp.ccminer.org' <86.202.176.224> port 3252
Setting DAG epoch #123...
 DCR: Stratum - Connected (yiimp.ccminer.org:3252)
 DCR: Authorized
Setting DAG epoch #123 for GPU1
Create GPU buffer for GPU1
Setting DAG epoch #123 for GPU0
Create GPU buffer for GPU0
Setting DAG epoch #123 for GPU2
Setting DAG epoch #123 for GPU3
Create GPU buffer for GPU3
Create GPU buffer for GPU2
GPU0 DAG creation time - 4984 ms
Setting DAG epoch #123 for GPU0 done
GPU1 DAG creation time - 5127 ms
Setting DAG epoch #123 for GPU1 done
GPU2 DAG creation time - 5125 ms
Setting DAG epoch #123 for GPU2 done
GPU3 DAG creation time - 5147 ms
Setting DAG epoch #123 for GPU3 done
ETH: 05/14/17-10:22:41 - New job from eu2.ethermine.org:4444
ETH - Total Speed: 95.401 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU0 23.872 Mh/s, GPU1 23.848 Mh/s, GPU2 23.870 Mh/s, GPU3 23.810 Mh/s
 DCR - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0
 DCR: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 0.000 Mh/s
 DCR: 05/14/17-10:22:46 - SHARE FOUND - (GPU 1)
 DCR: Share accepted (79 ms)!

[...] here the error occurs:

ETH - Total Speed: 94.658 Mh/s, Total Shares: 424, Rejected: 0, Time: 05:30
ETH: GPU0 23.246 Mh/s, GPU1 23.505 Mh/s, GPU2 23.848 Mh/s, GPU3 24.059 Mh/s
 DCR - Total Speed: 4354.281 Mh/s, Total Shares: 1122, Rejected: 8
 DCR: GPU0 1069.311 Mh/s, GPU1 1081.232 Mh/s, GPU2 1097.003 Mh/s, GPU3 1106.735 Mh/s
ETH: 05/14/17-10:05:18 - SHARE FOUND - (GPU 1)
ETH: Share accepted (40 ms)!
GPU0 t=63C fan=89%, GPU1 t=50C fan=87%, GPU2 t=57C fan=97%, GPU3 t=67C fan=90%
ETH: 05/14/17-10:05:36 - New job from eu2.ethermine.org:4444
ETH - Total Speed: 24.214 Mh/s, Total Shares: 425, Rejected: 0, Time: 05:31
ETH: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 24.214 Mh/s

[...]

ETH: 05/14/17-10:06:22 - SHARE FOUND - (GPU 3)
ETH: Share accepted (38 ms)!
ETH: 05/14/17-10:06:23 - New job from eu2.ethermine.org:4444
ETH - Total Speed: 24.262 Mh/s, Total Shares: 426, Rejected: 0, Time: 05:31
ETH: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 24.262 Mh/s
 DCR - Total Speed: 1116.045 Mh/s, Total Shares: 1122, Rejected: 8
 DCR: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 1116.045 Mh/s
GPU0 t=45C fan=80%, GPU1 t=32C fan=81%, GPU2 t=39C fan=82%, GPU3 t=67C fan=88%
WATCHDOG: GPU 0 hangs in OpenCL call, exit
WATCHDOG: GPU 0 hangs in OpenCL call, exit
WATCHDOG: GPU 1 hangs in OpenCL call, exit
WATCHDOG: GPU 1 hangs in OpenCL call, exit
WATCHDOG: GPU 2 hangs in OpenCL call, exit
WATCHDOG: GPU 2 hangs in OpenCL call, exit


▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ͻ
▒      Claymore's Dual ETH + DCR/SC/LBC/PASC GPU Miner v9.3      ▒
▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ͼ

ETH: 3 pools are specified
Main Ethereum pool is eu2.ethermine.org:4444
DCR: 2 pools are specified
Main Decred pool is yiimp.ccminer.org:3252

Catalyst 15.12 is REQUIRED for best performance and compatibility
For Polaris cards, use latest drivers
At least 16 GB of Virtual Memory is required for multi-GPU systems
Make sure you defined GPU_MAX_ALLOC_PERCENT 100
Be careful with overclocking, use default clocks for first tests
Press "s" for current statistics, "0".."9" to turn on/off cards, "r" to reload pools, "e" or "d" to select current pool
OpenCL initializing...

AMD Cards available: 4
GPU #0: Ellesmere, 1898 MB available, 36 compute units
GPU #0 recognized as Radeon RX 480/580
GPU #1: Ellesmere, 1900 MB available, 36 compute units
GPU #1 recognized as Radeon RX 480/580
GPU #2: Ellesmere, 1900 MB available, 36 compute units
GPU #2 recognized as Radeon RX 480/580
GPU #3: Ellesmere, 1900 MB available, 36 compute units
GPU #3 recognized as Radeon RX 480/580
POOL/SOLO version
GPU #0: algorithm ASM
GPU #1: algorithm ASM
GPU #2: algorithm ASM
GPU #3: algorithm ASM
No NVIDIA CUDA GPUs detected.
Total cards: 4
AMD ADL library not found.
ETH: Stratum - connecting to 'eu2.ethermine.org' <91.121.222.33> port 4444
ETH: Stratum - Connected (eu2.ethermine.org:4444)
DUAL MINING MODE ENABLED: ETHEREUM+DECRED
ETH: qtminer stratum mode
Watchdog enabled
Remote management (READ-ONLY MODE) is enabled on port 3333

ETH: Authorized
 DCR: Stratum - connecting to 'yiimp.ccminer.org' <86.202.176.224> port 3252
Setting DAG epoch #123...
 DCR: Stratum - Connected (yiimp.ccminer.org:3252)
 DCR: Authorized
Setting DAG epoch #123 for GPU2
Setting DAG epoch #123 for GPU0
Create GPU buffer for GPU0
Create GPU buffer for GPU2
GPU0 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.
Setting DAG epoch #123 for GPU1
Create GPU buffer for GPU1
GPU1 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.
GPU2 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.
Setting DAG epoch #123 for GPU3
Create GPU buffer for GPU3
GPU3 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.
ETH: 05/14/17-10:06:44 - New job from eu2.ethermine.org:4444
ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 0.000 Mh/s
 DCR - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0
 DCR: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 0.000 Mh/s
GPU 0 failed
GPU 1 failed
GPU 2 failed
Setting DAG epoch #123 for GPU2
GPU2, OpenCL error -38 - cannot write buffer for DAG
Setting DAG epoch #123 for GPU0
GPU0, OpenCL error -38 - cannot write buffer for DAG
Setting DAG epoch #123 for GPU1
Setting DAG epoch #123 for GPU3
GPU3, OpenCL error -38 - cannot write buffer for DAG
GPU1, OpenCL error -38 - cannot write buffer for DAG
GPU 3 failed
GPU 2 failed
GPU 0 failed
GPU 3 failed
GPU 1 failed
ETH: 05/14/17-10:06:59 - New job from eu2.ethermine.org:4444
ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 0.000 Mh/s
 DCR - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0
 DCR: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 0.000 Mh/s
GPU0 t=35C fan=69%, GPU1 t=29C fan=70%, GPU2 t=34C fan=71%, GPU3 t=37C fan=80%
WATCHDOG: GPU error, you need to restart miner

============ end console output ==================

[Alainmardirossian]

Hi

I was wondering if you solved your issue? Because i am having exactly the same problem with one of my rigs. My rig stops mining after 12 hours and it says cl error 61  cannot allocate big buffer for DAG. I need to reboot the rig to start minning for an other 12 hours. 

[LesPristy]

I have a similar problem here, it happened maybe an hour ago.

1xRX480 4GB
Ubuntu 16.04
Claymore dual-mining ETH+DCR

It's been running OK for a month. Then I received the error "cannot allocate big buffer for dag".

I didn't have to reboot, just restarted the claymore process and it started mining again. I hope it doesn't do it again because I would hate to have to babysit it.

[MATHReX]

Check you PAGE file size - it should be more than 16000 MBs. Keep minimum 16000 MBs and maximum 20000 MBs. A majority of the miners rely on virtual mem to keep the process quick and it requires a good amount of size allocation.

[vovan2007]

Hi!

I have the same problem
My configuration:
1 x R9 290 4GB
2 x R9 290x 4GB
4GB RAM and 2 GB SWAP

OS: Ubuntu 16.04.2
Drivers: AMDGPU-PRO 17.10
Software Claymore 9.3 (similary 9.4 and 9.5)

In my case, drop out only one graphics card randomly.

Code:

21:40:35:718    cd921740        args: -dcoin sia -dcri 15
21:40:35:735    cd921740
21:40:35:735    cd921740        ................................................................ͻ
21:40:35:735    cd921740        .      Claymore's Dual ETH + DCR/SC/LBC/PASC GPU Miner v9.3      .
21:40:35:735    cd921740        ................................................................ͼ
21:40:35:735    cd921740
21:40:35:966    cd921740        ETH: 2 pools are specified
21:40:35:966    cd921740        Main Ethereum pool is eu1.ethermine.org:4444
21:40:35:966    cd921740        SC: 2 pools are specified
21:40:35:966    cd921740        Main Siacoin pool is sia-eu1.nanopool.org:7777
21:40:38:269    cd921740        OpenCL platform: AMD Accelerated Parallel Processing
21:40:38:269    cd921740        OpenCL initializing...

21:40:38:269    cd921740        AMD Cards available: 3
21:40:38:269    cd921740        GPU #0: Hawaii, 4082 MB available, 44 compute units
21:40:38:269    cd921740        GPU #0 recognized as Radeon 290X
21:40:38:269    cd921740        GPU #1: Hawaii, 4082 MB available, 40 compute units
21:40:38:269    cd921740        GPU #1 recognized as Radeon 290
21:40:38:269    cd921740        GPU #2: Hawaii, 4082 MB available, 44 compute units
21:40:38:269    cd921740        GPU #2 recognized as Radeon 290X
21:40:38:269    cd921740        POOL/SOLO version
21:40:38:269    cd921740        b343
21:40:38:269    cd921740        Platform: Linux
21:40:38:471    cd921740        start building OpenCL program for GPU 0...
21:40:39:839    cd921740        done
21:40:40:043    cd921740        start building OpenCL program for GPU 1...
21:40:40:161    cd921740        done
21:40:40:365    cd921740        start building OpenCL program for GPU 2...
21:40:40:483    cd921740        done
21:40:40:691    cd921740        GPU #0: algorithm ASM
21:40:40:695    cd921740        GPU #1: algorithm ASM
21:40:40:701    cd921740        GPU #2: algorithm ASM
21:40:40:703    cd921740        cudaGetDeviceCount failed (35, CUDA driver version is insufficient for CUDA runtime version), probably no CUDA devices
21:40:40:703    cd921740        No NVIDIA CUDA GPUs detected.
21:40:40:703    cd921740        Total cards: 3
21:40:46:105    cd921740        AMD ADL library not found.
21:40:46:110    cd921740        amdgpu cnt: 3
21:40:46:110    cd921740        No NVIDIA cards in the list, NVML library will not be used.

#############################################################################

23:01:49:024    adffb700        ETH: 06/09/17-23:01:49 - New job from eu1.ethermine.org:4444
23:01:49:024    adffb700        target: 0x0000000112e0be82 (diff: 4000MH), epoch #128
23:01:49:024    adffb700        ETH - Total Speed: 78.727 Mh/s, Total Shares: 98, Rejected: 0, Time: 01:21
23:01:49:024    adffb700        ETH: GPU0 25.109 Mh/s, GPU1 26.396 Mh/s, GPU2 27.222 Mh/s
23:01:49:024    adffb700         SC - Total Speed: 1259.635 Mh/s, Total Shares: 50, Rejected: 0
23:01:49:024    adffb700         SC: GPU0 401.742 Mh/s, GPU1 422.336 Mh/s, GPU2 435.557 Mh/s
23:01:51:361    cd921740        GPU0 t=86C fan=97%, GPU1 t=81C fan=97%, GPU2 t=73C fan=97%
23:01:51:361    cd921740        em hbt: 1, dm hbt: 1, fm hbt: 40,
23:01:51:361    cd921740        watchdog - thread 0, hb time 22
23:01:51:361    cd921740        watchdog - thread 1, hb time 137
23:01:51:361    cd921740        watchdog - thread 2, hb time 138
23:01:51:361    cd921740        watchdog - thread 3, hb time 39
23:01:51:361    cd921740        watchdog - thread 4, hb time 1822
23:01:51:361    cd921740        watchdog - thread 5, hb time 1716        


23:02:12:433    adffb700        ETH: 06/09/17-23:02:12 - New job from eu1.ethermine.org:4444
23:02:12:433    adffb700        target: 0x0000000112e0be82 (diff: 4000MH), epoch #128
23:02:12:433    adffb700        ETH - Total Speed: 51.394 Mh/s, Total Shares: 98, Rejected: 0, Time: 01:21
23:02:12:433    adffb700        ETH: GPU0 25.001 Mh/s, GPU1 26.393 Mh/s, GPU2 0.000 Mh/s
23:02:12:433    adffb700         SC - Total Speed: 822.296 Mh/s, Total Shares: 50, Rejected: 0
23:02:12:433    adffb700         SC: GPU0 400.011 Mh/s, GPU1 422.285 Mh/s, GPU2 0.000 Mh/s
23:02:16:515    adffb700        ETH: checking pool connection...
23:02:16:515    adffb700        send: {"id":3,"jsonrpc":"2.0","method":"eth_getWork","params":[]}

23:02:16:574    adffb700        parse packet: 247
23:02:16:574    adffb700        ETH: job is the same
23:02:16:574    adffb700        new buf size: 0
23:02:21:394    cd921740        GPU0 t=86C fan=97%, GPU1 t=81C fan=97%, GPU2 t=59C fan=97%
23:02:21:394    cd921740        em hbt: 1, dm hbt: 1, fm hbt: 46,
23:02:21:394    cd921740        watchdog - thread 0, hb time 191
23:02:21:394    cd921740        watchdog - thread 1, hb time 75
23:02:21:394    cd921740        watchdog - thread 2, hb time 182
23:02:21:394    cd921740        watchdog - thread 3, hb time 82
23:02:21:394    cd921740        watchdog - thread 4, hb time 31855
23:02:21:394    cd921740        watchdog - thread 5, hb time 31749                             

23:02:51:425    cd921740        GPU0 t=86C fan=97%, GPU1 t=81C fan=97%, GPU2 t=52C fan=97%
23:02:51:425    cd921740        em hbt: 0, dm hbt: 1, fm hbt: 50,
23:02:51:426    cd921740        watchdog - thread 0, hb time 133
23:02:51:426    cd921740        watchdog - thread 1, hb time 19
23:02:51:426    cd921740        watchdog - thread 2, hb time 112
23:02:51:426    cd921740        watchdog - thread 3, hb time 13
23:02:51:426    cd921740        watchdog - thread 4, hb time 61886
23:02:51:426    cd921740        WATCHDOG: GPU 2 hangs in OpenCL call, exit
23:02:51:426    cd921740        watchdog - thread 5, hb time 61780
23:02:51:426    cd921740        WATCHDOG: GPU 2 hangs in OpenCL call, exit   

##################    After restart claymore miner  ################

Code:

23:02:54:566    302a6740        args: -dcoin sia -dcri 15
23:02:54:566    302a6740
23:02:54:566    302a6740        ................................................................ͻ
23:02:54:566    302a6740        .      Claymore's Dual ETH + DCR/SC/LBC/PASC GPU Miner v9.3      .
23:02:54:566    302a6740        ................................................................ͼ
23:02:54:566    302a6740
23:02:54:767    302a6740        ETH: 2 pools are specified
23:02:54:767    302a6740        Main Ethereum pool is eu1.ethermine.org:4444
23:02:54:767    302a6740        SC: 2 pools are specified
23:02:54:767    302a6740        Main Siacoin pool is sia-eu1.nanopool.org:7777
23:02:54:793    302a6740        OpenCL platform: AMD Accelerated Parallel Processing
23:02:54:793    302a6740        OpenCL initializing...

23:02:54:793    302a6740        AMD Cards available: 3
23:02:54:793    302a6740        GPU #0: Hawaii, 1949 MB available, 44 compute units
23:02:54:793    302a6740        GPU #1: Hawaii, 1949 MB available, 40 compute units
23:02:54:793    302a6740        GPU #2: Hawaii, 1949 MB available, 44 compute units
23:02:54:793    302a6740        POOL/SOLO version
23:02:54:793    302a6740        b343
23:02:54:793    302a6740        Platform: Linux
23:02:54:841    302a6740        start building OpenCL program for GPU 0...
23:02:54:933    302a6740        done
23:02:55:137    302a6740        start building OpenCL program for GPU 1...
23:02:55:230    302a6740        done
23:02:55:434    302a6740        start building OpenCL program for GPU 2...
23:02:55:525    302a6740        done
23:02:55:734    302a6740        GPU #0: algorithm ASM
23:02:55:738    302a6740        GPU #1: algorithm ASM
23:02:55:744    302a6740        GPU #2: algorithm ASM
23:02:55:746    302a6740        cudaGetDeviceCount failed (35, CUDA driver version is insufficient for CUDA runtime version), probably no CUDA devices
23:02:55:746    302a6740        No NVIDIA CUDA GPUs detected.
23:02:55:746    302a6740        Total cards: 3
23:03:05:198    302a6740        AMD ADL library not found.
23:03:05:203    302a6740        amdgpu cnt: 3
23:03:05:204    302a6740        No NVIDIA cards in the list, NVML library will not be used.
23:03:05:224    18a39700        ETH: Stratum - connecting to 'eu1.ethermine.org' <51.255.64.43> port 4444
23:03:05:294    18a39700        send: {"id":2,"jsonrpc":"2.0","method":"eth_login","params":["0x5DC7B5Bb80ac2D09FB7B9D230277FA7aF69f29BD.290x1","x"]}

23:03:05:294    18a39700        ETH: Stratum - Connected (eu1.ethermine.org:4444)
23:03:05:305    302a6740        DUAL MINING MODE ENABLED: ETHEREUM+SIACOIN

#############################################################################################################################


23:03:07:575    1923a700        Setting DAG epoch #128 for GPU2
23:03:07:575    1923a700        Create GPU buffer for GPU2
23:03:07:575    1ba3f700        Setting DAG epoch #128 for GPU0
23:03:07:575    1ba3f700        Create GPU buffer for GPU0
23:03:07:575    1923a700        GPU2 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.

23:03:07:575    1ba3f700        GPU0 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.

23:03:10:579    1aa3d700        GPU 1 failed
23:03:10:579    1a23c700        Setting DAG epoch #128 for GPU1
23:03:10:580    1a23c700        GPU1, OpenCL error -38 - cannot write buffer for DAG

23:03:10:580    19a3b700        Setting DAG epoch #128 for GPU2
23:03:10:580    19a3b700        GPU2, OpenCL error -38 - cannot write buffer for DAG

23:03:10:580    1b23e700        Setting DAG epoch #128 for GPU0
23:03:10:580    1b23e700        GPU0, OpenCL error -38 - cannot write buffer for DAG

23:03:10:580    1ba3f700        GPU 0 failed
23:03:10:580    1923a700        GPU 2 failed
23:03:13:580    1a23c700        GPU 1 failed
23:03:13:580    19a3b700        GPU 2 failed
23:03:13:580    1b23e700        GPU 0 failed      

After restarting rig everything works for one-two hours.

[Alainmardirossian]

Check you PAGE file size - it should be more than 16000 MBs. Keep minimum 16000 MBs and maximum 20000 MBs. A majority of the miners rely on virtual mem to keep the process quick and it requires a good amount of size allocation.

How can you do this on Ubuntu 16.04.2 ?

[Heye]

Did anyone find a solution? I'm having the same problem on linux and windows,
someone opened an issue on github about this https://github.com/nanopool/Claymore-Dual-Miner/issues/106

[hagbase]

Check you PAGE file size - it should be more than 16000 MBs. Keep minimum 16000 MBs and maximum 20000 MBs. A majority of the miners rely on virtual mem to keep the process quick and it requires a good amount of size allocation.

This is your solution. 

[Heye]

Check you PAGE file size - it should be more than 16000 MBs. Keep minimum 16000 MBs and maximum 20000 MBs. A majority of the miners rely on virtual mem to keep the process quick and it requires a good amount of size allocation.

This is your solution. 

I already tried. I can mine with xmr-stack just fine

[ninjabull]

I'm having what I'm guessing is the same issue but for CUDA.

I've tried increasing page file to 17-25GB, using the env. variables and even using the benchmark 10 flag. Nothing works, while etherminer is working fine.

I have never been able to run this miner. Any suggestions?

Code:

04:18:38:819	1ff0	Check and remove old log files...
04:18:38:824	1ff0	args: -benchmark 10 
04:18:38:827	1ff0	
04:18:38:831	1ff0	ÉÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍ»
04:18:38:835	1ff0	º     Claymore's Dual ETH + DCR/SC/LBC/PASC GPU Miner v10.2      º
04:18:38:838	1ff0	ÈÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍͼ
04:18:38:842	1ff0	
04:18:39:703	1ff0	OpenCL platform: NVIDIA CUDA
04:18:39:708	1ff0	AMD OpenCL platform not found 
04:18:40:451	1ff0	CUDA initializing...

04:18:40:456	1ff0	NVIDIA Cards available: 1 
04:18:40:461	1ff0	CUDA Driver Version/Runtime Version: 9.1/8.0
04:18:40:466	1ff0	GPU #0: GeForce GTX 970, 4096 MB available, 13 compute units, capability: 5.2

04:18:40:471	1ff0	Total cards: 1 
04:18:44:108	1ff0	No AMD cards in the list, ADL library will not be used.
04:18:44:114	1ff0	NVML version: 9.388.59
04:18:44:230	1ff0	CUDA error 0 #102 - 46
04:18:44:393	1ff0	CUDA error 0 #103 - 46
04:18:44:493	1ff0	CUDA error 0 #105 - 46
04:18:44:609	1ff0	

04:18:44:614	1ff0	
BENCHMARK MODE!

04:18:44:632	1e70	Setting DAG epoch #10...
04:18:45:075	970	ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
04:18:45:080	970	ETH: GPU0 0.000 Mh/s
04:18:45:683	1e70	Setting DAG epoch #10 for GPU0
04:18:45:688	1e70	Create GPU buffer for GPU0
04:18:45:774	1e70	CUDA error - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.

04:18:48:091	970	ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
04:18:48:091	970	ETH: GPU0 0.000 Mh/s
04:18:48:794	1e70	Set global fail flag, failed GPU0
04:18:48:794	1e70	GPU 0 failed
04:18:48:794	1e74	Setting DAG epoch #10 for GPU0
04:18:48:894	1e74	GPU 0, CUDA error 46 - cannot write buffer for DAG

04:18:51:110	970	ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
04:18:51:122	970	ETH: GPU0 0.000 Mh/s

Thank you.