Con Kolivas
|
b1b79e0efc
|
There is no need to hold the workbase lock when submitting a share
|
11 years ago |
Con Kolivas
|
50f7da96bb
|
Show the network or max pool diff along with the share diff in verbose logging
|
11 years ago |
Con Kolivas
|
d8882dacaf
|
Store the block height in proxy mode by extracting it from the coinbase
|
11 years ago |
Con Kolivas
|
29a8235c64
|
Only start the blockupdate thread if we're not in proxy mode
|
11 years ago |
Con Kolivas
|
09cf4f984e
|
Change the current workbase diff when the parent pool diff changes
|
11 years ago |
Con Kolivas
|
e36e5ce010
|
Submit the actual shares sent to the proxy upstream
|
11 years ago |
Con Kolivas
|
d265e51664
|
Create a send processing queue in the proxy generator and receive shares that meet the parent pool target
|
11 years ago |
Con Kolivas
|
72e6026052
|
Fix various coinbase assembly issues
|
11 years ago |
Con Kolivas
|
db1b2b45a4
|
Parse the mining difficulty when proxying and store it in the stratifier as the current workbase diff. Update notify and diff in stratifier when a change is deteced.
|
11 years ago |
Con Kolivas
|
747e689de9
|
Make each proxied client's enonce1 a function of the constant enonce1 and the extra bytes from the nonce2 which are unique for each client instance
|
11 years ago |
Con Kolivas
|
da383b8735
|
Store the notify data coming from the proxifier to the stratifier and create the basic work templates
|
11 years ago |
Con Kolivas
|
faf87ceeee
|
Store the proxy base subscribe details in a struct protected by the workbase lock
|
11 years ago |
Con Kolivas
|
fae9983fd7
|
Inform the stratifier process of the data in subscribe and notify from the generator proxy
|
11 years ago |
Con Kolivas
|
6ac33e57ac
|
Create initial stratum proxy connection
|
11 years ago |
Con Kolivas
|
78c1b2dfda
|
Make process instances an array within the ckpool struct allowing us to enable only those we need when we add more modules to future code
|
11 years ago |
Con Kolivas
|
8631f33809
|
Accept ping requests on the various unix sockets
|
11 years ago |
Con Kolivas
|
034e30ca8d
|
Don't try to send a message to clients that will be invalidated
|
11 years ago |
Con Kolivas
|
0e8fa71c11
|
Drop the connections to clients that are unauthorised, which also makes stratifier restarts work
|
11 years ago |
Con Kolivas
|
2ffdd791f7
|
Add support for dropclient being sent to the connector process
|
11 years ago |
Con Kolivas
|
93dafcf459
|
Store the ckpool struct pointer in workbase and client structs to avoid referencing the global pointer
|
11 years ago |
Con Kolivas
|
883c033ea3
|
Add share logging of whatever we can from rejected shares
|
11 years ago |
Con Kolivas
|
02438119b8
|
Show correct share count
|
11 years ago |
Con Kolivas
|
c05cbcd135
|
Fix warning
|
11 years ago |
Con Kolivas
|
78fbb4e91d
|
Make logging directory configurable, and store share logs by height/workid and totals for the pool as well
|
11 years ago |
Con Kolivas
|
4f106a6736
|
Display shares and share rate in logging
|
11 years ago |
Con Kolivas
|
3dc5f4c0c2
|
May as well store the share diff since we've calculated it
|
11 years ago |
Con Kolivas
|
5cfd5e7b17
|
Minor change
|
11 years ago |
Con Kolivas
|
c57519f4ff
|
Make job id a 64 bit variable to allow it to be unique on each restart, with time for the higher bits
|
11 years ago |
Con Kolivas
|
4b7af07036
|
Minor unlock cleanup
|
11 years ago |
Con Kolivas
|
a89e1fb9af
|
Add a stratum message sending function and a welcome message on connecting
|
11 years ago |
Con Kolivas
|
a524544367
|
Display hashrates also biased with time
|
11 years ago |
Con Kolivas
|
9c30d46b08
|
Bias early shares based on a simpler exponential function based on time since first share when calculating dsps
|
11 years ago |
Con Kolivas
|
b5caeedf73
|
Use a single rolling average for worker hashrate calculation and an exponential bias based on time to work out diff
|
11 years ago |
Con Kolivas
|
55045434f1
|
Make the start and minimum diff configurable and set them to 42 and 1 by default
|
11 years ago |
Con Kolivas
|
351e7470fb
|
Count rejected shares in diff management
|
11 years ago |
Con Kolivas
|
78abaedd2f
|
Add very basic share logging
|
11 years ago |
Con Kolivas
|
a615b56d76
|
Install signal handlers which push the signal to the top process which will try to gracefully shut down each child process
|
11 years ago |
Con Kolivas
|
764aa9717a
|
Use doubly linked lists for stratum messages
|
11 years ago |
Con Kolivas
|
457f1c17ee
|
Create a separate thread for sending messages to clients, not blocking on any clients not ready for reads
|
11 years ago |
Con Kolivas
|
c908088dc4
|
Add extended hour hashrate stats
|
11 years ago |
Con Kolivas
|
31021f3d13
|
Create a separate hashtable for disconnected clients and only store reusable instances, adding reusable stats
|
11 years ago |
Con Kolivas
|
eabd8848bc
|
Add basic pool stats, updating and displaying them in a separate thread every 15s
|
11 years ago |
Con Kolivas
|
d53715e139
|
Further tweak the diff, clamping it down more aggressively than up, store a 1min rolling average hashrate and display it once diff is stable
|
11 years ago |
Con Kolivas
|
7ef0915aea
|
Fix thinko in diff rate ratio calculation that was making it recalculate diff every time
|
11 years ago |
Con Kolivas
|
7389b7b15c
|
Cope with broken clients that don't send an integer for id
|
11 years ago |
Con Kolivas
|
8521e00151
|
Describe user associated with client id
|
11 years ago |
Con Kolivas
|
0e68d49b40
|
Only allow one connected instance per enonce1
|
11 years ago |
Con Kolivas
|
49b38997a0
|
Add more logging of share information at info level of logging
|
11 years ago |
Con Kolivas
|
9019dcb2a9
|
Fix deref on failed wb match
|
11 years ago |
Con Kolivas
|
e119e13fb1
|
Add share hashtable purging on new blocks
|
11 years ago |