Con Kolivas
|
6f7dc2d2f0
|
Add update detection in parse_node_msg and remove double warning
|
9 years ago |
|
Con Kolivas
|
5ddfcdb14e
|
Identify all node message types in the stratifier
|
9 years ago |
|
Con Kolivas
|
50c9b5ecd2
|
Move message type detection to stratifier
|
9 years ago |
|
Con Kolivas
|
c5718d510e
|
Detect node message type in the generator
|
9 years ago |
|
Con Kolivas
|
b09bb253fc
|
Act as a passthrough in node mode, passing what type of stratum message we're passing through
|
9 years ago |
|
Con Kolivas
|
eab56b032a
|
Set up local bitcoind connections in node mode
|
9 years ago |
|
Con Kolivas
|
8755e5c77b
|
Fix passthrough failures falling through as success
|
9 years ago |
|
Con Kolivas
|
0c605e3dee
|
Add basic node subscription functionality
|
9 years ago |
|
Con Kolivas
|
6a39cdb299
|
Add configuration option for node mode
|
9 years ago |
kanoi
|
51b892440b
|
ckdb - correct lock type in cmd_setopts
|
9 years ago |
|
Con Kolivas
|
85112ebce9
|
Flag a proxy as not alive when we are unable to send to it in passthrough mode
|
9 years ago |
|
Con Kolivas
|
2eaeb2a96c
|
Don't keep storing shares in redirector mode after we've attempted to redirect the client
|
9 years ago |
|
ckolivas
|
c437d3283b
|
Only try to redirect clients once in redirector mode, acting as a regular passthrough thereafter
|
9 years ago |
|
ckolivas
|
45a44ace7e
|
Fix passthrough/redirector trying to grab semaphore twice
|
9 years ago |
|
kanoi
|
bc1e08cd01
|
ckdb - remove duplicate case code for socket data
|
9 years ago |
|
kanoi
|
1c0f55c06d
|
ckdb - ensure the listener queues can empty
|
9 years ago |
|
kanoi
|
511d8b73a4
|
ckdb - itemise the sync values in 'homepage'
|
9 years ago |
|
kanoi
|
de3be722cd
|
ckdb - separate socket data processing
|
9 years ago |
|
kanoi
|
e6469e64c4
|
ckdb - fix recalculation of displayed reward for a shift
|
9 years ago |
|
kanoi
|
98cd4005dc
|
ckdb - comment ktree find before and after
|
9 years ago |
|
Con Kolivas
|
e1d6f62d3e
|
Name ckpassthrough if no name is given
|
9 years ago |
|
kanoi
|
a31e2555b4
|
ckdb - add a missing lock
|
9 years ago |
|
Con Kolivas
|
b6ceb9b484
|
Fix userproxies not having their semaphores initialised and avoid trying to process epoll responses from proxies not marked alive
|
9 years ago |
|
Con Kolivas
|
1e2e5bab7d
|
Add semaphores for subproxies
|
9 years ago |
|
Con Kolivas
|
0574a54554
|
Don't disable parent proxies in disable_subproxy for failover to work
|
9 years ago |
|
Con Kolivas
|
8ef853b8fc
|
Use the proxy connsock semaphore to serialise uses of the cs->buf to prevent races
|
9 years ago |
|
Con Kolivas
|
012d94dd35
|
Add reconnect sends to alive bool changing on proxy conditions to try and fail over/back
|
9 years ago |
|
Con Kolivas
|
6f842fc2d8
|
Cope with proxy auths that have no password
|
9 years ago |
|
Con Kolivas
|
96f055486c
|
We should serialise calls to GBT, not just update_base
|
9 years ago |
|
Con Kolivas
|
cab637c9b1
|
Use epoll for wait_write_select
|
9 years ago |
|
Con Kolivas
|
d2d5a5daaa
|
Don't use persistent connections to bitcoind
|
9 years ago |
|
Con Kolivas
|
83dbc9db12
|
Use epoll for wait_read_select
|
9 years ago |
|
Con Kolivas
|
11537481fd
|
Fail gracefully after 5 unsuccessful attempts to update_base
|
9 years ago |
|
Con Kolivas
|
06bdd7fa6a
|
Silence lack of current workbase warning in proxy mode since it happens till there is a proxy workbase to work with
|
9 years ago |
|
Con Kolivas
|
c7794a3d80
|
Retry getbase after a failure up to 5 times or indefinitely if a block change has been identified
|
9 years ago |
|
Con Kolivas
|
edd79bd8fa
|
Elaborate which bitcoind has failed responses
|
9 years ago |
|
Con Kolivas
|
432d1ce4d5
|
Don't treat POLLRDHUP as fatal unless we can't receive data after it, and return errors from wait_select functions when they close the socket
|
9 years ago |
|
Con Kolivas
|
c645a6fc69
|
Make the server watchdog a standalone thread that doesn't need messaging
|
9 years ago |
|
Con Kolivas
|
fba82db0f9
|
Typo
|
9 years ago |
|
Con Kolivas
|
61b25e4ac4
|
Add information about whether the attempt to reopen a socket was successful or not
|
9 years ago |
|
Con Kolivas
|
b5ea946ef4
|
Output generator started message as soon as we have the first live server
|
9 years ago |
|
Con Kolivas
|
7ce2afae83
|
Try to reopen any closed sockets in json_rpc_call
|
9 years ago |
|
Con Kolivas
|
b622a6ec8d
|
Check for errors in wait_read_select, closing the socket if they occur, and add a variant for non-unix sockets that checks for hangups as well
|
9 years ago |
|
Con Kolivas
|
6ff89b31a2
|
Check server fd is still valid as well as having its alive flag set
|
9 years ago |
|
Con Kolivas
|
bfbe3a1039
|
Don't block on dead servers if possible when calling server_alive from the gen_loop
|
9 years ago |
|
Con Kolivas
|
3fc984d7fe
|
Demote message
|
9 years ago |
|
Con Kolivas
|
8c474c1784
|
Support low diffs like those on testnet
|
9 years ago |
|
Con Kolivas
|
c1f812effc
|
Fix extremely unlikely race on fd being accessed before it is set to -1 on closing
|
9 years ago |
|
Con Kolivas
|
516a708aad
|
Fix bitcoind failover not working and keep connections open for faster switching in case of failure
|
9 years ago |
|
Con Kolivas
|
4aa73e29f8
|
Use the cs->fd handle directy in case it gets invalidated while we're waiting for a socket line
|
9 years ago |
