Frontier Mining Setup Notes (Ubuntu 15.04/Geth v1.0/AMD/Ethminer)

245

Comments

  • Cryptominer1Cryptominer1 Member Posts: 26
    Also set the GPU_MAX_ALLOC_PERCENT=95
  • sphinxsphinx Member Posts: 18
    I have a R9 270x 2GB card and using $ export GPU_MAX_ALLOC_PERCENT=95 and it still gives me GPU can't allocate the DAG in a single chunk. Bailing error?
  • mtbxmtbx Member Posts: 16
    I have a 7970 and am also running in the same issue on windows. Any ideas?

    ---
    Allocating/mapping single buffer failed with: clEnqueueWriteBuffer(-4).
    GPU can't allocate the DAG in a single chunk. Bailing.
    ---

    And I've set both variables to the following too

    SETX GPU_MAX_HEAP_SIZE 100
    SETX GPU_MAX_ALLOC_PERCENT 100
  • sphinxsphinx Member Posts: 18
    I think a single card R9 270x 2GB is not going to make the cut right?
  • xiaopowxiaopow Member Posts: 15
    edited August 2015
    @Cryptominer1, I didn't delete my old DAG, I set 'auto-dag' and 'no-precompute' and 'export GPU_MAX_ALLOC_PERCENT=95' and started mining again.

    @PhonikG @SilentCicero Thanks for the reply. If I setup another rig to mine, how do i connect the two rigs? Can I set them to deposit the mined ethers into the same account, how could you do that?
  • mikkaaamikkaaa Member Posts: 34
    $ export GPU_MAX_ALLOC_PERCENT=95
    $ ethminer -G --opencl-device 0 --no-precompute


    [OPENCL]:Allocating/mapping single buffer failed with: clCreateBuffer(-61). GPU can't allocate the DAG in a single chunk. Bailing.

    How to recover? And how about baking changes like these into the default launch values...
  • mikkaaamikkaaa Member Posts: 34
    edited August 2015
    Do I need to delete some old files, before these works?

    Edit. Miner works again - Reset everyhing and boot.
    Post edited by mikkaaa on
  • SilentCiceroSilentCicero Toronto, CAMember Posts: 159 ✭✭✭
  • Cryptominer1Cryptominer1 Member Posts: 26
    My rig only worked after export GPU_MAX_ALLOC_PERCENT=98 don't know why 95 didn't do the trick
  • BishopBishop Ontario, CanadaMember Posts: 2
    @SilentCicero thanks for the guide. Everything working tickety boo.
    @PhonikG thanks for the card. Had to GPU_MAX_ALLOC and everything running fine since 11:00am EST.

    Now I just have to get Teamviewer working with Ubuntu 15.04...... Minor detail I guess.
  • PhonikGPhonikG Member Posts: 41 ✭✭
    edited August 2015
    @Bishop careful with Teamviewer... I know of miners back in the day who had their pool changed, and If i remember correctly it had something to do with Teamviewer... you are always better off to use ssh with keys:

    https://help.ubuntu.com/community/SSH/OpenSSH/Keys

    ...and then use `screen` to keep things running after you log off.
  • unimerciounimercio Member Posts: 11
    @SilentCicero excellent post, thanks. My first mined block Just 4 hours later on a solo R9 280x success!
  • lfowlie890lfowlie890 Member Posts: 2
    @mikkaaa when you say "Reset everything" what do you mean? I'm getting the same error message with [OPENCL]:Allocating/mapping single buffer failed with: clCreateBuffer(-61). GPU can't allocate the DAG in a single chunk. Bailing.
  • boondoggerboondogger Member Posts: 7
    HELP! this guide was awesome....all went well except for last ethminer command..
    [email protected] ~> lspci | grep VGA
    01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Tahiti XT [Radeon HD 7970/8970 OEM / R9 280X]
    02:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Tahiti XT [Radeon HD 7970/8970 OEM / R9 280X]
    03:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Tahiti XT [Radeon HD 7970/8970 OEM / R9 280X]

    [email protected] ~>
    10:24 PM
    all cards still recognised by bios

    ethminer -G --opencl-device 0, or any of cards 1-3 same error below.
    (GETH SYNCED FINE)

    ERROR
    No GPU device with sufficient memory was found. Can't GPU mine. Remove the -G argument
    [email protected]:~$

    has anyone had this problem?
    any help appreciated.....

    cheers
    G
  • mikkaaamikkaaa Member Posts: 34

    @mikkaaa when you say "Reset everything" what do you mean? I'm getting the same error message with [OPENCL]:Allocating/mapping single buffer failed with: clCreateBuffer(-61). GPU can't allocate the DAG in a single chunk. Bailing.

    I guess the "GPU_MAX_ALLOC_PERCENT" only works after you reboot your computer. I didn't reinstall or delete any files manually..

  • boondoggerboondogger Member Posts: 7
    also can anyone point me to how we port olympic rewards to a frontier coinbase?
  • SilentCiceroSilentCicero Toronto, CAMember Posts: 159 ✭✭✭
  • dan123dan123 Member Posts: 7
    when i INSTALL i need to run this command-
    $ geth account new
    $ geth --rpc --rpccorsdomain "*" --genesis "~/genesis_block.json"
    or should i use localhost? like this
    $ geth account new
    $ geth --rpc --rpccorsdomain "localhost" --genesis "~/genesis_block.json"
    Please reply
    Thank you
  • SilentCiceroSilentCicero Toronto, CAMember Posts: 159 ✭✭✭
    @dan123 as of the Thaw you do not need to build and select the genesis block (i.e. "--genesis"). Genesis has been selected and has been baked into geth. Re: cors, `--rpccorsdomain "localhost"`. You need to also add in `--autodag` if I'm not mistaken. So Just
    $ geth --rpc --rpccorsdomain "localhost" --autodag
    should do.

    Please refer to this:
    https://blog.ethereum.org/2015/08/04/ethereum-protocol-update-1/,

    and this:
    https://blog.ethereum.org/2015/08/04/the-thawing-frontier/

    for more details on the Thaw and Genesis.
  • boondoggerboondogger Member Posts: 7
    thanks @SilentCicero, i like your method..

    finally
    all cards recoginised!

    used.
    $ export GPU_MAX_ALLOC_PERCENT=95
    (hit enter, no confirmation that anything happened)

    Great recipe...all goes well until final ethminer command.....beating my head....i used to use:

    ethminer -G -t 3 -F "http://internal IP:8545"

    but now attempt at doing 3 windows of:

    $ ethminer -G --opencl-device 0 --no-precompute
    $ ethminer -G --opencl-device 1 --no-precompute
    $ ethminer -G --opencl-device 2 --no-precompute

    gives me error on just trying the second window........

    [email protected]:~$ ethminer -G --opencl-device 0 --no-precompute
    [OPENCL]:Found suitable OpenCL device [Tahiti] with 2974810112 bytes of GPU memory
    miner 09:39:26|ethminer Getting work package...
    miner 09:39:26|ethminer Grabbing DAG for #290decd9…
    miner 09:39:30|ethminer Got work package:
    miner 09:39:30|ethminer Header-hash: 373683cddbca5e538ed55cc1b2d4a711600a5c4ff29c4198c70abb25e1d4d036
    miner 09:39:30|ethminer Seedhash: 290decd9548b62a8d60345a988386fc84ba6bc95484008f6362f93160ef3e563
    miner 09:39:30|ethminer Target: 0000000000bea44a2e9e9b93ceb5811ed9a597c486c5a92dc96aad1eaaa88064
    ℹ 09:39:30|gpuminer0 workLoop 0 #00000000… #290decd9…
    ℹ 09:39:30|gpuminer0 Initialising miner...
    [OPENCL]:Using platform: AMD Accelerated Parallel Processing
    [OPENCL]:Using device: Tahiti(OpenCL 1.2 AMD-APP (1445.5))
    miner 09:39:30|ethminer Mining on PoWhash #373683cd… : 0 H/s = 0 hashes / 0.5 s
    [OPENCL]:Printing program log
    [OPENCL]:LOOP UNROLL: pragma unroll (line 300)
    Unrolled as requested!
    LOOP UNROLL: pragma unroll (line 266)
    Unrolled as requested!
    LOOP UNROLL: pragma unroll (line 361)
    Unrolled as requested!
    LOOP UNROLL: pragma unroll (line 361)
    Unrolled as requested!

    [OPENCL]:Creating one big buffer for the DAG
    [OPENCL]:Allocating/mapping single buffer failed with: clCreateBuffer(-61). GPU can't allocate the DAG in a single chunk. Bailing.
    [OPENCL]:clEnqueueWriteBuffer(-38)

    i have three cards and can only mine on one......opening several windows and mining while geth is running gives me a bus error?

    to run geth i use:
    geth --autodag --etherbase "blah" --rpc console

    why wont single command (not multi ethminer windows)
    ethminer -G -t 3 -F "http://localhost:8545" --no-precompute

    could there an issue with running slots on gen1 in bios?

    I know iam missing something simple...

    has anyone seen this before?
    any help appreciated.
    thanks!
  • dan123dan123 Member Posts: 7
    so from what i understand, from the guide, this-
    3. Download/build the genesis block
    $ cd ~
    $ curl -O https://raw.githubusercontent.com/ethereum/genesis_block_generator/master/mk_genesis_block.py
    $ curl -O https://bootstrap.pypa.io/get-pip.py
    $ sudo python get-pip.py
    $ sudo pip install bitcoin
    $ sudo pip install --upgrade bitcoin
    $ python mk_genesis_block.py --extradata 0x11bbe8db4e347b4e8c937c1c8370e4b5ed33adb3db69cbdb7a38e1e50b1b82fa > genesis_block.json
    Is not needed any more?

    And then, on the next command-
    4. Create an account (note, use a strong passphrase and that this RPC cors setting is unsafe), then start geth, let it load the genesis block and sync with the network, watch for the block numbers (e.g. #24179)
    $ geth account new
    $ geth --rpc --rpccorsdomain "*" --genesis "~/genesis_block.json"
    We do instead-
    $ geth account new
    $ geth --rpc --rpccorsdomain "localhost" --autodag
    Is that correct?
    Thank you.
  • heaveancornerheaveancorner Member Posts: 27
    Hello all. I'm Chris.
    thank for this awesome tutorial. (Sorry for my basic English)
    After 2 days, I'm mining ethereum at least with one card !!
    Now i try to mine with 2 cards (7950's) and my hashrate is splitted between the cards ! I can not find the problem. My hashrate is 20 Mhs with one card or 10 Mhs by cards !!

    Do you think it is because of the DAG ? how to delete old file and create new one for each cards?

    I'm using --auto-dag with geth and also the command "export GPU_MAX_ALLOC_PERCENT=95".

    I'm not skilled it is already an exploit haha. Thank you if you can help me.
  • SilentCiceroSilentCicero Toronto, CAMember Posts: 159 ✭✭✭
    @dan123 yep, that should work.
  • SilentCiceroSilentCicero Toronto, CAMember Posts: 159 ✭✭✭
    @heaveancorner are you mining with each card in their own terminal/instance?

    i.e.

    1st Command Line Terminal:
    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 0 --no-precompute
    2nd Command Line Terminal:
    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 1 --no-precompute
    if not, I would try this.
  • SilentCiceroSilentCicero Toronto, CAMember Posts: 159 ✭✭✭
    edited August 2015
    @boondogger if you havent already, try each card in their own terminal, like the example I posted above, where GPU max allocation is set before each instance of ethminer. Also for geth, make sure you set the '--rpccorsdomain "localhost"'.

    How much RAM do you have in your machine? I would recommend at least 4gb of RAM for multi-card/GPU mining.

    If there was issues with the motherboard configuration or BIOS that would likely lead to a crash or shutdown (at least in my experience).

    This issue looks to be more a software issue related to ethminer/GPU_MAX settings. You may also want to try setting 'GPU_MAX_ALLOC_PERCENT' to 98, i.e. (`export GPU_MAX_ALLOC_PERCENT=98`) before running each instance of ethminer.

    So I would try something like this:
    $ export GPU_MAX_ALLOC_PERCENT=98
    $ ethminer -G --opencl-device 1 --no-precompute
    Let me know how it goes.
  • heaveancornerheaveancorner Member Posts: 27

    @heaveancorner are you mining with each card in their own terminal/instance?

    i.e.

    1st Command Line Terminal:

    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 0 --no-precompute
    2nd Command Line Terminal:
    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 1 --no-precompute
    if not, I would try this.
    i'm trying now.
    One terminal is already mining (21MHS) GPU 0
    The other terminal is generating the dag file GPU 1

    Will see. I guess i'll have an error at the end.

    system 4gb RAM + AMD A6 + MSI A75A-g3s + 7950 oc edition
  • heaveancornerheaveancorner Member Posts: 27

    @heaveancorner are you mining with each card in their own terminal/instance?

    i.e.

    1st Command Line Terminal:

    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 0 --no-precompute
    2nd Command Line Terminal:
    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 1 --no-precompute
    if not, I would try this.
    i'm trying now.
    One terminal is already mining (21MHS) GPU 0
    The other terminal is generating the dag file GPU 1

    Will see. I guess i'll have an error at the end.

    system 4gb RAM + AMD A6 + MSI A75A-g3s + 7950 oc edition
    Same problem ... when the second card start mining, first gpu drop from 20 MHs to 10 and the other instance is 10 MHs too.

    Both instance looks like copy (same powhash, same speed ..)

    Strange
  • heaveancornerheaveancorner Member Posts: 27

    @heaveancorner are you mining with each card in their own terminal/instance?

    i.e.

    1st Command Line Terminal:

    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 0 --no-precompute
    2nd Command Line Terminal:
    $ export GPU_MAX_ALLOC_PERCENT=95 
    $ ethminer -G --opencl-device 1 --no-precompute
    if not, I would try this.
    i'm trying now.
    One terminal is already mining (21MHS) GPU 0
    The other terminal is generating the dag file GPU 1

    Will see. I guess i'll have an error at the end.

    system 4gb RAM + AMD A6 + MSI A75A-g3s + 7950 oc edition
    Same problem ... when the second card start mining, first gpu drop from 20 MHs to 10 and the other instance is 10 MHs too.

    Both instance looks like copy (same powhash, same speed ..)

    Strange
    I need to find the DAG file or reset/delete something.

  • heaveancornerheaveancorner Member Posts: 27
    one card is cold so I guess she is not running ...
  • SilentCiceroSilentCicero Toronto, CAMember Posts: 159 ✭✭✭
    @heaveancorner what are your PSU specs? - also, did you try this:
    $ export GPU_MAX_ALLOC_PERCENT=98 
    $ ethminer -G --opencl-device 0 --no-precompute

    $ export GPU_MAX_ALLOC_PERCENT=98
    $ ethminer -G --opencl-device 1 --no-precompute
Sign In or Register to comment.