Con Kolivas
618b3e8b11
Track queued and staged per pool once again for future use.
12 years ago
Con Kolivas
4ca288e820
Limit queued_getworks to double the expected queued maximum rather than factoring in number of pools.
12 years ago
Con Kolivas
ad90269508
Minimise the number of getwork threads we generate.
12 years ago
Con Kolivas
0feb679b67
Only keep the last 6 blocks in the uthash database to keep memory usage constant. Storing more is unhelpful anyway.
12 years ago
Con Kolivas
b74b54d95b
Check we haven't staged work while waiting for a curl entry before proceeding.
12 years ago
Con Kolivas
61df3013a8
Ignore the submit_fail flag when deciding whether to recruit more curls or not since we have upper bounds on how many curls can be recruited, this test is redundant and can lead to problems.
12 years ago
ckolivas
edd9b81622
Do not add time to dynamic opencl calculations over a getwork.
12 years ago
Con Kolivas
9de3a264fc
Increase max curls to number of mining threads + queue * 2, accounting for up and downstream comms.
12 years ago
Con Kolivas
3ab5dba67e
Queue enough requests to get started.
12 years ago
Con Kolivas
3ebe8e8c77
Revert "Scale maximum number of curls up according to work submission rate."
...
This reverts commit 1dff48e759
.
Did not increase staged work even though it decreased queues.
12 years ago
Con Kolivas
3ceb57b8f6
There is no point trying to clone_work in get_work() any more since we clone on every get_work_thread where possible.
12 years ago
Con Kolivas
787e40a7cc
There is no point subtracting 1 from maxq in get_work_thread.
12 years ago
Con Kolivas
1dff48e759
Scale maximum number of curls up according to work submission rate.
12 years ago
Con Kolivas
56be75228e
Roll back to 45f0ac7b48
12 years ago
Con Kolivas
c1886c9c99
Assume we need a full allotment of work after lp and make clone_work clone_lpwork only.
12 years ago
Con Kolivas
82fa6e25d5
There is no point running through clone_work from get_work() now that cloning is done from the get_work_thread.
12 years ago
Con Kolivas
caa266ca3b
Make opt_queue a function of mining threads in clone_work as well.
12 years ago
Con Kolivas
652efc6939
opt_queue should be a function of the number of mining threads.
12 years ago
Con Kolivas
8ed381f78a
Only set lagging flag once there are no staged work items.
12 years ago
Con Kolivas
f5ac84b804
select_pool does not switch back to the primary once lagging is disabled.
12 years ago
ckolivas
f71d635f0f
Free work before retrying in get_work_thread.
12 years ago
Con Kolivas
b3cfe47222
Increment total work counter under mutex lock.
12 years ago
Con Kolivas
e8daf1d8f2
Increment the queued count after the curl is popped in case there's a delay waiting on curls and we think we've queued work when in fact we're waiting on curls.
12 years ago
Kano
b3e9858c9f
API new command 'coin' with mining information
12 years ago
Con Kolivas
08948e02f3
Increase curl reaping time to 5 minutes since comms between curl requests can be 2 mins apart with lots of rolltime.
12 years ago
Con Kolivas
51d485e236
No need for extra variable in hash_push.
12 years ago
Con Kolivas
c01284fb07
Remove short options -r and -R to allow them to be reused and remove readme entries for deprecated options.
12 years ago
Con Kolivas
8fc701dd76
Deprecate the opt_fail_pause parameter, leaving a null placeholder for existing configurations.
12 years ago
Con Kolivas
35cb7fd101
Don't pause after failed getwork, set lagging flag and reassess.
12 years ago
Con Kolivas
69d8c339aa
Add message to share if it's a resubmit.
12 years ago
Con Kolivas
b622e97e3e
We should not be pausing in trying to resubmit shares.
12 years ago
Con Kolivas
a9278ce944
Get rid of the extending fail pause on failed connects since we discard work after a period.
12 years ago
Con Kolivas
34b8dcb7bc
get_work always returns true so turn it into a void function.
12 years ago
Con Kolivas
260b1c87c0
get_work never returns false so get rid of fail pause loop.
12 years ago
Con Kolivas
4df9c387f0
Get rid of pause and retry from get_upstream_work so we only do it from one place.
12 years ago
Con Kolivas
1f295cd7fa
Deprecate the opt_retries feature as no one wants cgminer to automatically abort. Leave a null placeholder for configurations that still have it.
12 years ago
Con Kolivas
f6d2853c9b
Summary layout difference.
12 years ago
Con Kolivas
ed3eb5771a
Introduce a new statistic, Work Utility, which is the number of difficulty 1 shares solved per minute. This is useful for measuring a relative rate of work that is independent of reject rate and target difficulty.
12 years ago
Con Kolivas
26ab7346e3
Implement a new pool strategy, BALANCE, which monitors work performed per pool as a rolling average every 10 minutes to try and distribute work evenly over all the pools.
...
Do this by monitoring diff1 solutions to allow different difficulty target pools to be treated equally, along with solo mining.
Update the documentation to describe this strategy and more accurately describe the load-balance one.
12 years ago
Con Kolivas
d8c76bbd08
Getwork fail was not being detected. Remove a vast amount of unused variables and functions used in the old queue request mechanism and redefine the getfail testing.
12 years ago
Con Kolivas
8fa708093b
Don't try to start devices that don't support scrypt when scrypt mining.
12 years ago
Con Kolivas
b814b42c7f
Consider us lagging only once our queue is almost full and no staged work.
12 years ago
Con Kolivas
b60f9da4a5
Simplify the enough work algorithm dramatically.
12 years ago
Con Kolivas
0e56dffd92
Only queue from backup pools once we have nothing staged.
12 years ago
Con Kolivas
e16b7566bd
Don't keep queueing work indefinitely if we're in opt failover mode.
12 years ago
Con Kolivas
6d1949e694
Make sure we don't opt out of queueing more work if all the queued work is from one pool.
12 years ago
Con Kolivas
a5f7b953fe
Set lagging flag if we're on the last of our staged items.
12 years ago
Con Kolivas
ee9e621e1c
Reinstate clone on grabbing work.
12 years ago
Con Kolivas
55f7c14982
Grab clones from hashlist wherever possible first.
12 years ago
Con Kolivas
c3e32274ee
Cull all the early queue requests since we request every time work is popped now.
12 years ago