I think I am now 100% sure that there is no switch to the primary pool from the fall-back pool not 30 min later, not 10 (which I set in --set-reconnection-timer 10), not at all.
So, I gave it a check and it worked as it should. What did I do for a try? Well I uses this config: --server eu.minexpool.nl,optiminer.eu
--port 9998,9999
--user XQZoBeGPZJoLhTdkcF8bScsNH5Wyesd9g4.W033,XQZoBeGPZJoLhTdkcF8bScsNH5Wyesd9g4.W033
--pass x,x
--max-connection-attempts 5
--set-reconnection-timer 10
--set-short-stats-interval 30
--set-long-stats-interval 120
So you see I reduced stats output a bit and set the reconnection timer to 10 minutes. And then got this output: Stratum connection not reacting for too long time.
Lost connection to stratum server eu.minexpool.nl:9998 or server not reachable.
Trying to connect in 5 seconds
Stratum connection not reacting for too long time.
Lost connection to stratum server eu.minexpool.nl:9998 or server not reachable.
Trying to connect in 5 seconds
Stratum connection not reacting for too long time.
Lost connection to stratum server eu.minexpool.nl:9998 or server not reachable.
Trying to connect in 5 seconds
Stratum connection not reacting for too long time.
Lost connection to stratum server eu.minexpool.nl:9998 or server not reachable.
Trying to connect in 5 seconds
Stratum connection not reacting for too long time.
Lost connection to stratum server eu.minexpool.nl:9998 or server not reachable.
Trying to connect in 5 seconds
Too many attempts. Switching to failover pool optiminer.eu:9999
Connected to optiminer.eu:9999
Subscribed to stratum server
New target received: 0000333333333333340000000000000000000000000000000000000000000000
New job received: cf6f
Authorized worker: XQZoBeGPZJoLhTdkcF8bScsNH5Wyesd9g4.W033
Start Mining...
[Device 0] Average speed (30s): 1211.59 sol/s
Submitting share
Share accepted
New job received: cf70
Submitting share
Share accepted
[Device 0] Average speed (30s): 1203.9 sol/s
[Device 0] Average speed (30s): 1159.75 sol/s
New job received: cf71
---------------------------------------------
Average speed (120s): 1180.03 sol/s
---------------------------------------------
[Device 0] Average speed (30s): 1143.1 sol/s
[Device 0] Average speed (30s): 1128.57 sol/s
[Device 0] Average speed (30s): 1082.13 sol/s
[Device 0] Average speed (30s): 1108.65 sol/s
Submitting share
Share accepted
---------------------------------------------
Average speed (120s): 1112.11 sol/s
---------------------------------------------
[Device 0] Average speed (30s): 1128.52 sol/s
[Device 0] Average speed (30s): 1100.85 sol/s
New target received: 0001051eb854a875700000000000000000000000000000000000000000000000
New job received: cf72
New job received: cf73
[Device 0] Average speed (30s): 1124.74 sol/s
Submitting share
Share accepted
[Device 0] Average speed (30s): 1114.01 sol/s
Submitting share
Share accepted
---------------------------------------------
Average speed (120s): 1113.44 sol/s
---------------------------------------------
[Device 0] Average speed (30s): 1112.48 sol/s
[Device 0] Average speed (30s): 1113.21 sol/s
[Device 0] Average speed (30s): 1106.15 sol/s
Submitting share
Share accepted
[Device 0] Average speed (30s): 1142.23 sol/s
---------------------------------------------
Average speed (120s): 1121.17 sol/s
---------------------------------------------
[Device 0] Average speed (30s): 1121.42 sol/s
[Device 0] Average speed (30s): 1100.78 sol/s
[Device 0] Average speed (30s): 1121.37 sol/s
[Device 0] Average speed (30s): 1139.7 sol/s
Trying to reconnect to original pool.
Lost connection to stratum server optiminer.eu:9999 or server not reachable.
Trying to connect in 5 seconds
Stratum connection not reacting for too long time.
Lost connection to stratum server eu.minexpool.nl:9998 or server not reachable.
Trying to connect in 5 seconds
Too many attempts. Switching to failover pool optiminer.eu:9999
Connected to optiminer.eu:9999
Subscribed to stratum server
New target received: 0000333333333333340000000000000000000000000000000000000000000000
New job received: cf73
Authorized worker: XQZoBeGPZJoLhTdkcF8bScsNH5Wyesd9g4.W033
[Device 0] Average speed (30s): 1102.12 sol/s
[Device 0] Average speed (30s): 1107.46 sol/s
New job received: cf74
New job received: cf75
[Device 0] Average speed (30s): 1068.84 sol/s
You see that of cause the first connection to minexpool.nl failed because I configured the wrong port. After 5 attempts it connected to optiminer.eu and started mining. After about 10 minutes it tried to reconnect to minexpool.nl, but failed again (because the port is still not valid) and therefore swaped back to the first failover. Overall this is exactly what it should do. Test system is an AMD A10 7850K integrated graphics, running Windows 10 Pro. |
|
|
|
I think I am now 100% sure that there is no switch to the primary pool from the fall-back pool not 30 min later, not 10 (which I set in --set-reconnection-timer 10), not at all.
I will give that a check. [Device 0] Average speed (5s): 9188.96 sol/s
[Device 1] Average speed (5s): 9918.42 sol/s
[Device 2] Average speed (5s): 10140.2 sol/s
[Device 3] Average speed (5s): 10173.8 sol/s
[Device 4] Average speed (5s): 9941.01 sol/s
[Device 5] Average speed (5s): 10024.8 sol/s
[Device 6] Average speed (5s): 6902.02 sol/s
[Device 7] Average speed (5s): 6636.27 sol/s
[Device 8] Average speed (5s): 6775.64 sol/s
[Device 9] Average speed (5s): 6401.12 sol/s
[Device 10] Average speed (5s): 7138.37 sol/s
[Device 11] Average speed (5s): 6737.25 sol/s
[Device 12] Average speed (5s): 6953.81 sol/s
Maybe a overall high CPU load? This seems to me that the runtime can not fill the cards fast enough with new work. What can help is increasing --set-work-batch. I hope to get is less demanding on CPU side timing in next major version. |
|
|
|
Interesting... you decompiled or extracted my code I see ... so we get Nvidia update for your miner soon? To open it up: my OpenCL code uses the Blake2B routine from older ZCash miner zogminer. You can find the code here: https://github.com/nginnever/zogminer/blob/master/src/libzogminer/kernels/cl_zogminer_kernel.clThere you also find the comment snipped you cited. This miner is under MIT license, so I think I am fine to use it. But you are right the nheqminer has same comment in, but is CUDA and not OpenCL. Since you should not be only one to see it, here is the full blake2b code of lolMiner 0.33: void round0(ulong8 blake_state, __global uint4 *results, __local uint *counter, uint tid, uint gId) {
if (tid > 26214) return;
ulong v[16];
// shift "i" to occupy the high 32 bits of the second ulong word in the
// message block. The lower will be used to modify the nounce
ulong word1 = ((ulong)tid << 32) | gId;
// init vector v
v[0] = blake_state.s0;
v[1] = blake_state.s1;
v[2] = blake_state.s2;
v[3] = blake_state.s3;
v[4] = blake_state.s4;
v[5] = blake_state.s5;
v[6] = blake_state.s6;
v[7] = blake_state.s7;
v[8] = blake_iv[0];
v[9] = blake_iv[1];
v[10] = blake_iv[2];
v[11] = blake_iv[3];
v[12] = blake_iv[4];
v[13] = blake_iv[5];
v[14] = blake_iv[6];
v[15] = blake_iv[7];
// gFunc in length of data
v[12] ^= 144 /* length of "i" */;
// last block
v[14] ^= (ulong)-1;
// round 1
gFunc(v[0], v[4], v[8], v[12], 0, word1);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 2
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], word1, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 3
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, word1);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 4
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, word1);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 5
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, word1);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 6
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], word1, 0);
// round 7
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], word1, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 8
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, word1);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 9
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], word1, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 10
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], word1, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 11
gFunc(v[0], v[4], v[8], v[12], 0, word1);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], 0, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
// round 12
gFunc(v[0], v[4], v[8], v[12], 0, 0);
gFunc(v[1], v[5], v[9], v[13], 0, 0);
gFunc(v[2], v[6], v[10], v[14], 0, 0);
gFunc(v[3], v[7], v[11], v[15], 0, 0);
gFunc(v[0], v[5], v[10], v[15], word1, 0);
gFunc(v[1], v[6], v[11], v[12], 0, 0);
gFunc(v[2], v[7], v[8], v[13], 0, 0);
gFunc(v[3], v[4], v[9], v[14], 0, 0);
v[0] = v[0] ^ blake_state.s0 ^ v[8];
v[1] = v[1] ^ blake_state.s1 ^ v[9];
v[2] = v[2] ^ blake_state.s2 ^ v[10];
v[3] = v[3] ^ blake_state.s3 ^ v[11];
v[4] = v[4] ^ blake_state.s4 ^ v[12];
v[5] = v[5] ^ blake_state.s5 ^ v[13];
v[6] = v[6] ^ blake_state.s6 ^ v[14];
v[7] = v[7] ^ blake_state.s7 ^ v[15];
uint4 output;
uint bucket; uint pos;
output.s0 = 5*tid;
output.s1 = (v[0] & 0xFFFFFFFF);
output.s2 = (v[0] >> 32);
output.s3 = (v[1] & 0xFFFFFFFF);
getBucket(bucket, output.s1);
getRem(output.s1, output.s1);
pos = atomic_inc(&counter[bucket]);
if (pos < bSize) results[bucket*bSize + pos] = output;
output.s0 = 5*tid+1;
output.s1 = (v[1] >> 32);
output.s2 = (v[2] & 0xFFFFFFFF);
output.s3 = (v[2] >> 32);
getBucket(bucket, output.s1);
getRem(output.s1, output.s1);
pos = atomic_inc(&counter[bucket]);
if (pos < bSize) results[bucket*bSize + pos] = output;
output.s0 = 5*tid+2;
output.s1 = (v[3] & 0xFFFFFFFF);
output.s2 = (v[3] >> 32);
output.s3 = (v[4] & 0xFFFFFFFF);
if (output.s0 < 131072) {
getBucket(bucket, output.s1);
getRem(output.s1, output.s1);
pos = atomic_inc(&counter[bucket]);
if (pos < bSize) results[bucket*bSize + pos] = output;
}
output.s0 = 5*tid+3;
output.s1 = (v[4] >> 32);
output.s2 = (v[5] & 0xFFFFFFFF);
output.s3 = (v[5] >> 32);
if (output.s0 < 131072) {
getBucket(bucket, output.s1);
getRem(output.s1, output.s1);
pos = atomic_inc(&counter[bucket]);
if (pos < bSize) results[bucket*bSize + pos] = output;
}
output.s0 = 5*tid+4;
output.s1 = (v[6] & 0xFFFFFFFF);
output.s2 = (v[6] >> 32);
output.s3 = (v[7] & 0xFFFFFFFF);
if (output.s0 < 131072) {
getBucket(bucket, output.s1);
getRem(output.s1, output.s1);
pos = atomic_inc(&counter[bucket]);
if (pos < bSize) results[bucket*bSize + pos] = output;
}
}
I hope you see that this is taken from Zogminer but modified for Minex. Also nheqminer has a similar phrase in also under MIT. To other sources I did not look. By the way, how do you like my round code? That one is due to my own and I like its simplicity  Question is now what should I do with you. I do not like binary kernels as you use because they make the miners incompatible, i.e. I could not support Vega with binary kernels. But if you found a simple way of extracting my OpenCL text form code I must come up with some more hiding... Was this really nessesary? |
|
|
|
@Lolliedieb
I suspect there is a big problem above v0.22, probably v0.31..v0.33
You miss one point: the nethash displayed by the pools is taken from the deamon connected to the pool (since every pool runs a full minex node so it can add blocks). This deamon calculates the nethash from the average time it took to mine the last blocks and the current network difficulty. So if this calculation says 200Msols it is likely that the true speed is between 180 and 220M. From the numbers I see in my fee connectors the 0.3x versions together do about 150Msols and all 0.2x versions about 30M... so it is almost impossible that all 0.3x versions only send wrong shares that can not build a block. Especially I have seen blocks build by my fee connectors (approx 3 to 4 a day) so this versions do propper calculations. Finally when pool sends a message that share is accepted and the pool is displaying sol/s I would expect the pool had checked the share it gets. If you claim this happens on your pool but the pool is not able to build a block from accepted shares something might be horribly wrong with the pool. Overall I think you just had bad luck (with 0.3x) and good one with 0.22. At those rates the volatility is very high. @Global diff: the recent down in prize and high diff made the profitability sink below level of other coins so miners left. The numbers I see here fit, so the sol/s arriving fit to nethash calculated from block time and my market share. I think all is ok here. |
|
|
|
That is nice speed on 980 Ti,
on 1080 i got 19.8 kSol/s on stock speeds, on PL 85: around 19.6 kSol/s.
I'm also curious what are speed compered to linux version of this miner?
Somewhare I saw that 1080 can reach 23 kSol/s on linux, and I was unable to even past 20 kSol/s on Windows.
Well that could be water cooled then or in fact a 1080TI - usually Windows and Linux should not make a difference. But it is questionable if such OC is worth because the consumption is high then too. I usually recommend running lolMiner at 75 to 85 % power target. fall back pool does not seem to work. minexpool is out for half a day and the miner does not switch.. or it takes so long that hiveos just turns it off
how long is the connection delay? please reduce it
The connection delay is 5 seconds and it tries to reconnect 5 times per default ... so after half a minute it should switch. I also saw many miners switch, but for some it did not work. What I also saw is that the 0.30 to 0.32 versions have problems when the primary pool is not available when the miner starts - then they may just crash. That was resolved in 0.33. But I will double check if there are general problems with this feature that needs fix in next version. By the way: Please look at your rigs if these are correctly configured.I see one rig with about 100k sol mining to my account with the 0.31 example configuration - not that I have something against donations, but I think this is not intended  |
|
|
|
The --exit-on-zero option doesn't seem to be working, I just had my rig run for about half hour at 0 before I caught it...
I will try it if its working here. Did you disable long stats? Because when that print out is disabled its not workkng. v0.31 gives better hashrates for me than v0.33
I second that, especially on low end GPU (1050Ti) Did you try --optim OFF? Usually then the 1050 should do better on 0.33 - well but I may be wrong. Hello Lolliedieb.
Is there an option to lower the GPU load?
When using minor on the main PC, I want to run a miner while doing other work.
Not yet, but will put this to the list of requested features. Optiminer is better than lolminer Hashrate different 1000 sol/s
For fiji at the moment yes, can confirm that is hat problems with lolMiner. I will try to fix it in 0.4. Please help me. 13 GTX 1060 but only 6 GTX Hash~ 9k sol @@. i want all GTX 1060 Hash ~ 9k Sol
An other thing where I thinks its because of the current thread handling - also here I hope to make this happen less often in 0.4. ----------------------- To set an end to the other discussion: There was a certain demand for a Nvidia optimized and for a Windows port, too. I have to admit that now with some time behind me a kind of soft start would have been better - first do the more stable Linux variant and later on the Windows cross-build. When 0.1 came out the Windows build compiled first time just the day before... But anyways, I think this discussion is not new. When optiminer came out I guess the situation was similar, but with Windows miner crying because their profitability on mxminer was gone and they got no proper alternative. Same situation but a few month ago. I believe for the maturity of a coins infrastructure the support of as many platforms as possible is mandatory. As for the optiminer vs lolMiner part: I still think optiminer did a great job - I myself mined a bit with his miner in the beginning and it worked quite well. On the first versions some stability problems, but also that got better. And the algorithm is still better for AMD miners - the two mining softwares have approx the same speed, but his version is less memory hungry (works equally well on 4g and 8g AMD cards) and most rigs consume less energy on his miner - as said on same speed level. Only thing I still do not like much is accusing me for not paying attention to the GPL license - without any indicator what makes him / her thinking this. I went through all my code last days and checked back the licenses of code that I did not wrote. Its all either license free (I think English word is public domain) or MIT license, so I think the closed source distribution of the miner is fine. Still you have to believe my words, so here is my offer. As far as I know you are also located in middle Europe - correct me if I am wrong. If you like I invite you to my home (lets clarify details in private message), so you can have an exclusive look into my sources and convince yourself that everything is fine. If you have to stay over night because of distance I will pay your accommodation or prepare our guest room for you. Be my guest. |
|
|
|
Under what license is this miner distributed?
Is the source available? It seems to use GPLed code, so I would expect it to be.
Hi there, Can you be more specific which code under GPL you think I am using? The only external libraries I use and link at the moment are the boost library and the g++/mingw64 runtimes. Well and OpenCL of cause. I hope I am not wrong here, so if you found somethink I use that I am not allowed to I will remove it immidiately. I wrote most of this from scratch - don't even have the external equihashverify code from node code in here. Beeing a bit more specific could help clarifying this. Anyone working with mining rig 6x GTX 1080ti on Windows 10 pro??
I have 4 rigs with 1080ti and 2 rigs with 1080 .. all with 6 cards.
With 1080 it is working without any problem.
But on any my 1080ti rig it not working. After some minutes rig unexpected restarted (no BSOD dump file generated). When I set device only on 1,2,3,4,5 (except 0) it works about 20-40 minutes when it crash. With device 0 active it will crash in 2-5 minutes.
Any tips? I have new updates and new drivers too. I tried to reset OC to default (i tried set less OC then default too), set powerlimit to 80 or 90 or 100 or 110% but nothing helped.
Thanks for any help
Sorry to hear that but a bit hard to debug from the distance. Could be too much CPU load or work batch turned toi high or well - just a runtime bug of the Windows binaries (not that unlikely - there still seem to be certain of them  ) Have you tried running the rig with hiveos or any other Linux? If its crashing there too its likely that this due to system stability - if not its my shitty Windows binaries... I am a bit sirry that these are not better yet :/ |
|
|
|
Well ... there is an extra forum here where you can announce and discuss new pools - maybe setting then a link in the Minex general pool is a good idea. But here this is getting a bit to much if I get one pool advertising per page - this is for discussing and giving support for lolMiner. I have to announce that I just released version 0.33.The reason is that I investigated the bug that some users claim the miner disconnecting from pool. I do not have a full log here but just by inspecting the code I found a problem that may have caused this and where the miner defiantly did not handle things like it should be. Well - I do not know if this causes the problem, but I fixed at least one issue now that I did not want to keep in code until 0.4 is done Link is - as usual - on the thread front page. Also I now computed the checksums of the lolMiner binaries so you can compare if files you get are integer. |
|
|
|
...
[Device 0] Average speed (5s): 8374.28 sol/s
[Device 1] Average speed (5s): 8435.19 sol/s
[Device 2] Average speed (5s): 8431 sol/s
[Device 3] Average speed (5s): 8510.09 sol/s
[Device 4] Average speed (5s): 8481.73 sol/s
...
Sorry, that snap was a bit too short To do some debugging I need to know in which state the stratum connection was and what counters were running / or were long expired at the time your connection stops working. So to get a complete picture of the situation I require the log from approx the moment where you got the last difficulty / new job adjustment (but connection was working fine till then) until you decided to stop the miner because its not sending shares. That should be long enough so I can see which of my safety timers (that should prevent this) is not working. The --max-connection-attempts will only work if you define a failover pool and then still the miner has to detect that the connection is broken. Usually it should detect this at some time and then reconnect, but somehow it does not do so. Therefore would be great having the longer log and the config. (But plz in private message - I guess this could be a loooooong post here else ) What we can expect from 0.4?  0.08 more than 0.32?  Thats for sure ^^ I currently plan to do some changes in the work scheduling. I aim for making the stuff less dependent on the memory on the GPUs so the 4G cards should do same speed as the 8G (on AMD in particular), since this currently is not the case. Also I want to exclude the status message code into an external module. Currently there are only two large code fragments, one for mining and one for the connection. This will be connection, mining and logging then. Pulling out the stats code out of the mining section will allow me to do more fancy stuff with it, e.g. providing a JSON API for the miner, so you can grab statistics for external analysis. One further point I have on my list - but do not know yet if that will be 0.4 or later is a re-design of my mining code itself. Currently AMDs are suffering from the code to eat vector registers when compiling like Candy preventing the AMD scheduler to do efficiently place work on the card and cover all latencies. The re-structuring there should also allow to introduce more switchable optimizations. Yea, I think thats it |
|
|
|
[Device 0] Average speed (5s): 12551.9 sol/s
[Device 1] Average speed (5s): 7562.11 sol/s
[Device 2] Average speed (5s): 7417.15 sol/s
[Device 3] Average speed (5s): 12431.9 sol/s
I have seen this only twice yet - on Ubuntu systems - but could be the same reasons. Can you have a look how much your system load is? It seems this happens when the miner does not get enough CPU time to load all cards with their task all of the time. One solution could be to manually set the priority of the miner to "higher" - this solved it for the one guy. The second case was in my own testing system - a 1080 starting at 19k sol and then after some time dropping to 14.5k while the 2nd card still working at high rate. Here the slower card went so hot that it was thermal throttling. Open the case and allowing higher fan speed solved this. Whats the output of nv-smi when your cards are running? |
|
|
|
hi. I have a question for you. Is it possible to make the --optim and --set-work-batch parameter set for each video card separately? This is necessary if there are different video cards.
At the moment this won't work, but for the next version it will likely be. For work batch I can guaranty it, with optim I am not sure because I will do some rework of the OpenCL code and therefore the exact way optim will work is not yet fixed Mine is stable with 4gpu's i remove the other gpu in shift to eth algo, I hope the 8gpu limit will resolve soon by the @Dev
As you may have read in the previous page: There is no limit by lolMiner. It can use all cards it gets reported by the OpenCL drivers without any internal restriction, therefore there is nothing I can fix. When running Windows you must have recent Windows 10 version and recent GPU drivers. Additionally if you have AMD GPUs it seems that the blockchain drivers have problems with more then 8 GPUs. On the regular AMD Crimson drivers you still have to set the GPU profile to "Compute", here is a manual how it is done: http://www.aurelp.com/2017/11/01/solved-how-to-add-more-than-8-gpus-in-windows-10-step-by-stepfix/The newest version 0.32beta has a problem staying connected to pool. After a few hours (totaly random) miner continues hashing (and spending electricity) but worker is not connected to the pool (not mining for me).
Why so bold? No need to highlight your text that way - I am reading every post here! Can you send me the console log of your miner when this happens? Usually when the connection does down I should be able to see something in this logs. Also miner should reconnect or something like this when a share is send but it doesn't get an answer from pool. So either something is broken on pool side or one of my many safety mechanisms is not working properly (e.g. Reconnect to pool when no new job for 10 minutes, reconnect when no response to send share for 10 seconds, miner should stop sending work to GPU when there is no connection on the user connector (when fee one is broken it should get slower but not stop working) ... ) You see there are certain things that should prevent this, but I can not say why this does not work, so your config file and the text log of the miner in such a situation would be great so I can hunt down this bug. |
|
|
|
That's nice, how you do that? Are you using the v.32? Can you please share with me your .bat file config? I am using 13 rx 570 cards.. I just want to try the latest version, I think I miss something in my bat file..
Thanks dude!!
It seems that you must allow Windows 10 to update to recent build version and then reinstall current graphic dribers: https://bitcointalk.org/index.php?topic=2302268.0 |
|
|
|
Hi all, see first post in this thread, 0.32 is out Only smaller changes this time, but in particular since we were suffering from some mining pools under attack the last week the update may be welcome, since it fixes the problems the miner had on startup. I will now concentrate on 0.4 Have fun mining. |
|
|
|
In one of previous posts you mentioned that this week you will offer bounty for watchdog script for the Windows miner. Do you still plan to do that or you have changed your mind?
I already have one here that I still have to review - also must agree with the creator on the exact usage terms. On the other hand I have the feeling that doing something based on an API will be much more mature and that currently the solution with the .bat file I ship with the miner is not that bad, when deactivating the Windows GUI messages as described by some user a few pages earlier here in this thread. Therefore no news on this yet, since I have to clean up my mind for this. @Danik777: Reported to a moderator... I don't want such stuff here... It only works if I use the option: --set-work-batch 2G
It makes 7400 sols/s, no overclocking, Windows 7/64.
It can do more then 9k, but for that higher work batch will be needed at the moment. I do not understand why the AMD windows drivers restrict some users - and others not. Thats why the default is 2G at the moment for Windows. What you could try is running two command lines with the miner open, both mining on the same card. This should have similar effect as a higher work batch combined. I will redesign the work scheduling for 0.4 to ensure the algorithm can handle the speed of higher work batches without being so memory hungry. Hello guys. I`m trying to run 3 GPU`s and the CPU together with one bat file but i failed. When run the CPU with second bat file, aside from all gpu`s, all works but the hashrate on the cards drops alot. Did i make something wrong ? When i put "--platform 0" Start oly the CPU, when change to "1" start the gpu`s without the CPU
Your config looks fine, but remember that the CPU is required to schedule work to the graphic cards and also to test their solutions. I just can not recommend running the miner at the CPU, because it will slow down your graphic cards and a CPU at the current difficulty is likely to burn more energy then it mines as coin reward. Great to hear that! You are doing a great job Tried this MxMiner, got a total of 1,7 kH from my rig vs your lolminer 50 kH Thanks By the way we will likely having 0.32 today - only a bugfix release without to much news, but it fixes something that was bugging us last days multiple times. Thus when its out I can concentrate on doing 0.4 |
|
|
|
Is it really required to advertise in huge letters in a thread you did not open? I resist this time for reporting to Moderator, but will do next time. There are other opportunities enough like doing a proper post here and putting this message to your bottom line. Also, is there a way to auto-restart the miner after some time has passed? E.g you set the timer so the miner restarts itself once every hour.
Please make API as all the other miners. A few specialized mining operating systems would like to integrate your miner, and that is impossible at the moment because no hashing data (hasning speed, log in general) is coming out.
For all having massive stability problems, especially with Windows I'd like to point you to hive os. That mining OS is linux based and supports lolMiner from last version. From what I heard till now the integrated watchdogs work just perfect with the miner and systems are very stable. Give it a try! Also gives nice stats overview over your rigs. But of cause next version will come (with few corrections soon, with much more massive changes later). Stability is still a point I put a lot of effort in making it better and I promise we will have some API - not on 0.32, but maybe in 0.4 and for sure in 0.5. |
|
|
|
It seems the current build of 0.31 has a problem some times when connecting to other pools then suprnova and minexpool.nl - this not happens always, but some times and only when starting the miner. Once it is running fine it continues to run. The same happens to 0.22 by the way but in a more extreme version (that is what marlinkanamo just posted). I am currently searching the exact problem and will hopefully find a fix soon. Currently only "fix" is running 0.21 when this appears - the version does not seem to be affected, sorry for that |
|
|
|
No hash information is available in the mine pool
Do I need to modify anything?
It seems that your pool does not send an authorization message so the miner knows the worker name you put in is valid. The protocol behavior (that suprnova, minexpool, optiminer, 4rigs and co are using) that is expected by the miner equals the one of zcash: https://github.com/slushpool/poclbm-zcash/wiki/Stratum-protocol-changes-for-ZCashyou see after the miner subscribed to the pool it sends a mining.authorize request, but your pool does not answer ({"id": 2, "result": true, "error": null}\n for success, insert false if not). Therefore it does not start mining. |
|
|
|
My final verdict is, this miner can only run 8gpu's as i tested it in my 10gpu rig, the miner only detect 8gpu, not worth it right now to run in 10gpu rig... I hope the dev will test or fix the miner's problem of not detecting 10gpu..
Well we had running more then 12 cards with this miner. But if this works hardly depends on drivers and OS. For example Nvidia drivers on Windows 10 does not allow more then 8 cards to work, while the drivers for Linux do. But thats not my fault then, the miner can use all OpenCL devices the drivers report - there is no hard coded limitation on this. Problems with reconnecting to the pool https://mnx.4rigs.cf. An error occurs during the authorization phase - the miner exits. The program starts and stops indefinitely. Can just confirm this. I am hunting this down right now. How do you get that speed in your rx 480? mine is rx 480 8gb and i only got 8k sol/s, is any special setting to achieve that hash speed? can you please share your setting, I only use the default clock and mem, I only adjusted the voltage in AB @ -55 and -10 powlim
Current the AMDs need --work-batch as high as possible and --optim STD to max out their performance. Nvidias are much less sensitive to work batch changes. Will see where this comes from. |
|
|
|
How can I set temp limit in this miner?
This is plainly not possible - use power target in your driver to monitor abd set temperatures. I have the same issue trying to run a 4x 1080ti rig stock, except mine just keeps running at 0 (maybe I killed it before it crashed though), no errors, just goes from 20k sols/card to 0.
Once and a while it will run for a bit but will crash if it looses connection (not sure why it lost connection, all the other rigs on that switch are still hashing away with no problems).
I had many reports that the 0 sols appear more often when increasing OC. lolMiner utilizes and stresses card a lot and 0 sols is an indicator that the card failed in completing its task properly and the miner stops there because the complete event is never called. Try to start at clock 0, power target 75% and 0 mem and then increase slowly mem when you see the config is stable. deskop remote pls
demo claymore
Can you describe more in detail what feat you like to have? |
|
|
|
i have same "problem", but I'm lucky because i dont want to use rx 580... So same rig with amd and nvidia, and it only detects nvidia..
--list-devices do not work. It keeps runing command prompt again and again, over and over (shut down then start, shut down then start). but miner works fine, when start mining and only detecting nvidia..
A quick cite from the lolMiner_readme.pdf: --platform <number>, --device <device numbers>, --list-devices
OpenCL is organized in platforms, that correspond to the driver to be used for the Miner. On many system –
in particular when the CPU features an integrated graphic card – there is more then one platform on the
system available.
The parameter platform can be used to select the right platform. lolMiner offers the option --list-devices.
When lolMiner is started with this parameter it will print out a list with all platforms that were found on the
system. Additionally for every platform all devices are printed that are found within a platform.
After printing out the miner will close directly. To see the output in Windows you must run cmd (the
Windows command line), navigate to the directory where lolMiner-mnx.exe is in and then run lolMiner-
mnx.exe --list-devices directly from the command line. Writing the --list-devices to the configuration file is
not supported and running in the batch will end up in the miner closing instantaneously and you do not see a
print out.
From lolMiner-mnx 0.3 there is one special value AUTO that can be passed to the miner and also is the
default. This one will select the platform with the most graphic cards automatically. If there are two with the
same number, e.g. and Intel integrated GPU and one AMD / Nvidia graphic card it will prefer the non-Intel
platform.
So ... is not so hard is it? when you do not pass anything it seems the AUTO option gets the Nvidia platform... so you have to search and select the AMD by hand. P.S. I hope the stability of Windows improves...
I hope so, too ^^ Waiting for new Ubuntu to come out - then I can set up new Windows build chain with updates runtime libraries - hope that this helps improving stability. |
|
|
|
|
Show Posts
