Pages:
Author

Topic: DGB DigiByte Qubit and Myriad-Groestl Suprnova Pools - page 3. (Read 10013 times)

legendary
Activity: 1510
Merit: 1003
Seems that it helps, thanks!
legendary
Activity: 2688
Merit: 1240
what's the hell with the myr-groestl pool last hour? It cannot be just a bad luck ((

huh ? check the round shares, we're at 57% of the current estimated shares, it's all good
check last hour! It is 227% luck, block 4539386 was with 1143% luck. With such low block time as for DGB it cannot be bad luck

Last three blocks are well below 50%..

I've checked the logs and cannot see any errors but i've restarted the daemon for you
legendary
Activity: 1510
Merit: 1003
what's the hell with the myr-groestl pool last hour? It cannot be just a bad luck ((

huh ? check the round shares, we're at 57% of the current estimated shares, it's all good
check last hour! It is 227% luck, block 4539386 was with 1143% luck. With such low block time as for DGB it cannot be bad luck
legendary
Activity: 2688
Merit: 1240
what's the hell with the myr-groestl pool last hour? It cannot be just a bad luck ((

huh ? check the round shares, we're at 57% of the current estimated shares, it's all good
legendary
Activity: 1510
Merit: 1003
what's the hell with the myr-groestl pool last hour? It cannot be just a bad luck ((
legendary
Activity: 2688
Merit: 1240
So which version of ccminer should be used? I'm using the one from theblocksfactory (v1.2 with myr-gr) since the dashboard shows the correct hash rate.

ccminer 2.0 reports a better hash rate than with 1.2 but the pool doesn't agree - I'm getting ~30% of what ccminer shows me.

I'm mining with a GTX1070 on Win10 64b.

Best,
mksm

 The tpruvot ccminer 2.0 is what I've been using, shows up in theblocksfactory pool hashrate as quite close to the miner reported hashrate if you let it settle down for 10-15 perhaps 20 at most minutes.

I've got stuff running on that test machine that's going to take some time to finish, will go test out the changes to suprnova in an hour or so.

 (edit)
 confirm d= is working but d=0.03 in my testing wasn't giving very many blocks and 30-50% of "rated" hashrate range.
 d=0.02 is working quite a bit better so far, actually seeing a pool-side hashrate over 55 Mh/s (vs miner at 86+) for the first time (current rate has flattened out at a little over 70 MH/s and seems inclined to stay there give or take small normal variations).

 VARDIF however is STILL bouncing up to crazy high values then taking forever to come back down again - no visible change.

 VARDIF will consistantly bounce my miner up to a value in the .1-.12 difficulty range on it's first adjustment that happens something like a minute after the start - which is VERY high for this algo and yields very few shares to work on - then the SECOND adjustment 30 seconds to a minute later consistantly bounces the diff up OVER 0.5 which yields pretty close to ZERO blocks - and then it takes a long time to get back down UNDER 0.1 where a significant number of the shares are at on difficulty.
 When I watch the actual difficulty reported for each share, I see pretty close to none of them at 0.1 or higher, very few at 0.05 or higher, and the bulk of them at under 0.03



Yup, just use the static difficulty setting, that's why it's implemented
legendary
Activity: 1498
Merit: 1030
So which version of ccminer should be used? I'm using the one from theblocksfactory (v1.2 with myr-gr) since the dashboard shows the correct hash rate.

ccminer 2.0 reports a better hash rate than with 1.2 but the pool doesn't agree - I'm getting ~30% of what ccminer shows me.

I'm mining with a GTX1070 on Win10 64b.

Best,
mksm

 The tpruvot ccminer 2.0 is what I've been using, shows up in theblocksfactory pool hashrate as quite close to the miner reported hashrate if you let it settle down for 10-15 perhaps 20 at most minutes.

I've got stuff running on that test machine that's going to take some time to finish, will go test out the changes to suprnova in an hour or so.

 (edit)
 confirm d= is working but d=0.03 in my testing wasn't giving very many blocks and 30-50% of "rated" hashrate range.
 d=0.02 is working quite a bit better so far, actually seeing a pool-side hashrate over 55 Mh/s (vs miner at 86+) for the first time (current rate has flattened out at a little over 70 MH/s and seems inclined to stay there give or take small normal variations).

 VARDIF however is STILL bouncing up to crazy high values then taking forever to come back down again - no visible change.

 VARDIF will consistantly bounce my miner up to a value in the .1-.12 difficulty range on it's first adjustment that happens something like a minute after the start - which is VERY high for this algo and yields very few shares to work on - then the SECOND adjustment 30 seconds to a minute later consistantly bounces the diff up OVER 0.5 which yields pretty close to ZERO blocks - and then it takes a long time to get back down UNDER 0.1 where a significant number of the shares are at on difficulty.
 When I watch the actual difficulty reported for each share, I see pretty close to none of them at 0.1 or higher, very few at 0.05 or higher, and the bulk of them at under 0.03

newbie
Activity: 4
Merit: 0
Use ccminer 2.0 tpruvot...

Yeah, that's the one from: https://github.com/tpruvot/ccminer/. Shows a much lower hash rate in the dashboard.
hero member
Activity: 963
Merit: 500
This would be a cmdline using a static diff of 0.03:

Code:
./ccminer -a myr-gr -o stratum+tcp://dgbg.suprnova.cc:7978 -u suprnova.1 -p d=0.03

Thats working just perfect... thanks
also sent some donations  Wink


So which version of ccminer should be used? I'm using the one from theblocksfactory (v1.2 with myr-gr) since the dashboard shows the correct hash rate.

ccminer 2.0 reports a better hash rate than with 1.2 but the pool doesn't agree - I'm getting ~30% of what ccminer shows me.

I'm mining with a GTX1070 on Win10 64b.

Best,
mksm

Use ccminer 2.0 tpruvot...
newbie
Activity: 4
Merit: 0
So which version of ccminer should be used? I'm using the one from theblocksfactory (v1.2 with myr-gr) since the dashboard shows the correct hash rate.

ccminer 2.0 reports a better hash rate than with 1.2 but the pool doesn't agree - I'm getting ~30% of what ccminer shows me.

I'm mining with a GTX1070 on Win10 64b.

Best,
mksm
legendary
Activity: 2688
Merit: 1240
This would be a cmdline using a static diff of 0.03:

Code:
./ccminer -a myr-gr -o stratum+tcp://dgbg.suprnova.cc:7978 -u suprnova.1 -p d=0.03
legendary
Activity: 2688
Merit: 1240
I've tweaked the Vardiff settings and also enabled manual diff setting via the password now, try again !
legendary
Activity: 1498
Merit: 1030
On the DGB-Groestl pool the issue is definitely the difficulty - the VARDIF kicks it up to insane levels VERY quickly and then takes forever to move back down at all when it DOES move back down - if one is lucky one gets about 1 MINUTE of mining as something resembling full rate before the VARDIF goes crazy, usually on it's second adjustment.

The difficulty modifiers in ccminer do NOT help at all on this issue.

 GTX 1080 get hit VERY bad by this, but 1070s get hit by it pretty hard too.

 IMO kick the max VARDIF down to around .1 difficulty (as reported by ccminer) if you want this pool to be useable for higher-end cards.


You can either: Just wait for 2-3 mins after starting mining to let vardiff settle a bit.

Or set a direct difficulty by using d=4 as password, whereas 4 is the diff you want.


 The VARDIF difficulty never settles at a resonable level - it STAYS crazy high for a lot longer than 2-3 minutes, when running multi-card 1070 or even on my single-card 1080 rig - I've seen it STILL being crazy-high and pretty close to ZERO shares being generated after well over an hour of mining.
 It's AFTER the 2-3 minute period that it starts going crazy-high - the first couple of minutes of mining work well THEN the difficulty gets jumped up so high you get pretty close to zero shares.
 It does NOT "settle down some" after 2-3 minutes - if anything, it settles UP way too bloody high to be usable starting at that point.
 Even after an HOUR of mining I was rarely seeing even 50% of my MINER-side hashrate on the pool, more commonly 5%-20%, because the diff was so high no shares were getting generated, and often I'd see 3-5 blocks go by with ZERO shares at all.

 This was all using the same ccminer settings (except pool-specific ones) that I had been using on theblocksfactory's DGB-Groestl pool and getting 90% or better average of "miner hashrate" over a 5+ minute period on the poolside there (I know to expect variation).

 I tried doing the "set difficulty using d=4 as a password" thing, didn't seem to do anything with ANY of the settings I tried on that option - SAME identical behavior when I was testing that idea on my 1080 single rig.
 It also was NOT documented on that pool but I remembered seeing it as an option on at least one of your other pools so I figured I'd try it.


 (Edit)
 Yep, dashboard looks much better now, it actually has information on it!
 8-)


 The difficulty the pool reports on the dashboard page isn't even close the difficulty ccminer is assigning - not sure if that's an issue or if it's more of a "different reporting" type thing.

 d=# fixed difficulty option DEFINITELY is not working on the groestl pool - just tried it again and I'm seeing it do VARDIF difficulty increases.
 VARDIF difficulty actually DOES drop now, though it's slow to do so - as I recall from my earlier testing it would go up but didn't seem to EVER want to go back down again even after 5-10 minutes of getting almost no shares and watching multiple "new block" messages per share.

 It doesn't drop very fast though, and it keeps jumping WAY back up as soon as I start getting more than 1 or 2 shares per block again - pool reported hashrate PEAKED at about 50% of my miner hashrate, and has been DROPPING since about the first 2 minutes - it spent 3-4 minutes getting LESS THAN 3 Mh/sec ON THE POOL on a 85+ Mh at the miner single GTX 1080 rig, ballpark 15 minutes into this test run.
 It seems to be going back up some now, but it's still never seen 50% of miner rate.

 I can't say what exactly is wrong, but VARDIF on this pool is DEFINITELY not working right or even CLOSE to right - my guess is that it's set to aim at a target that is WAY TOO HIGH.

 I'm going to leave the one rig pointed this way overnight just to see if it makes a difference - but after looking at my stats from when I had ALL of my 1070 rigs and the 1080 pointed at this pool, I note that the highest full hour I had (out of about 4 hours I had my machines pointed this way) was around HALF of the "miner-reported" hashrate of all of them combined, which indicates that the VARDIF too high issue does NOT "settle down" at any point.


 (second edit)

 Checked suprnova GAME pool - that's working reasonably well, 5-10% variation from one hour to the next and getting about the same hashrate reported by the pool as I used to see out of LiteCoinPool or LiteGuardian when I was mining Litecoin.

 Hour and a half into the current test run, on DGB-Groestl pool giving me 8 Mh/sec down to ZERO over period covered by the "dashboard", and less than 30 Mh/sec on the "full hour" report in the middle of this run - on an 85+ Mh (miner side) rig that pulls fairly close to that on theblocksfactory.
 VARDIF does NOT "settle in" at any point at a REASONABLE difficulty level when looking at the miner itself as far as I can scroll it back.
 
 Forget this pool 'till the issue(s) get FIXED - VARDIF is definitely BROKEN on this DGB-Groestl pool, going way too high and STAYING there for way too long at a time.

legendary
Activity: 2688
Merit: 1240
On the DGB-Groestl pool the issue is definitely the difficulty - the VARDIF kicks it up to insane levels VERY quickly and then takes forever to move back down at all when it DOES move back down - if one is lucky one gets about 1 MINUTE of mining as something resembling full rate before the VARDIF goes crazy, usually on it's second adjustment.

The difficulty modifiers in ccminer do NOT help at all on this issue.

 GTX 1080 get hit VERY bad by this, but 1070s get hit by it pretty hard too.

 IMO kick the max VARDIF down to around .1 difficulty (as reported by ccminer) if you want this pool to be useable for higher-end cards.


You can either: Just wait for 2-3 mins after starting mining to let vardiff settle a bit.

Or set a direct difficulty by using d=4 as password, whereas 4 is the diff you want.


 
 I also noticed that the difficulty numbers reported by the POOL seem to have ZERO relationship to the difficulty numbers reported by the miner - this might be related to the core issue.
 I also noticed that the "dashboard" page on this pool doesn't seem to be working well if at all, as ALL of the graphs never report anything on that page.


I've fixed the dashboard, thanks for info.



 The only reason I even looked at this pool is that theblocksfactory's DGB-Groestl pool seems to have crashed and it taking a while to get back up, or it's getting DDOSed (though to be fair it seems like that pool shorts me a significant amount of hashrate too).





That was an issue with the dashboard, fixed.
legendary
Activity: 1498
Merit: 1030
 On the DGB-Groestl pool the issue is definitely the difficulty - the VARDIF kicks it up to insane levels VERY quickly and then takes forever to move back down at all when it DOES move back down - if one is lucky one gets about 1 MINUTE of mining as something resembling full rate before the VARDIF goes crazy, usually on it's second adjustment.

The difficulty modifiers in ccminer do NOT help at all on this issue.

 GTX 1080 get hit VERY bad by this, but 1070s get hit by it pretty hard too.

 IMO kick the max VARDIF down to around .1 difficulty (as reported by ccminer) if you want this pool to be useable for higher-end cards.

 
 I also noticed that the difficulty numbers reported by the POOL seem to have ZERO relationship to the difficulty numbers reported by the miner - this might be related to the core issue.
 I also noticed that the "dashboard" page on this pool doesn't seem to be working well if at all, as ALL of the graphs never report anything on that page.



 The only reason I even looked at this pool is that theblocksfactory's DGB-Groestl pool seems to have crashed and it taking a while to get back up, or it's getting DDOSed (though to be fair it seems like that pool shorts me a significant amount of hashrate too).


newbie
Activity: 8
Merit: 0
Hi Ocminer,
I m using ccminer 2.0 on an gtx 1080 with myr-gr algorithm on dgbg.suprnova.cc.

I ve mined for few hours yesterday, and also the results are very poor ( about 30% coins productivity ). On other pools still not 100% but close to 80%. 


Your pool , after about 20 minutes shows at max less then half hashrate. On other topics i found that may be something wrong with the difficulty settings from the pool.
I tried on ccminer -f 2 , -m 2 , and other numbers  as diff  multipliers but still no improvement.

The week before i have used sgminer on few amd cards, works well about 80% from expected.

Right now i m on other pool , but i ll be glad to come back, once is fixed .

Thanks.
legendary
Activity: 2688
Merit: 1240
Hey folks,

I've added two Pools for DGB:

Qubit:
https://dgbq.suprnova.cc

Myriad-Groestl:
https://dgbg.suprnova.cc

Happy hashing !
Pages:
Jump to: