Inaba is correct about the database access needing multi-threading, or to be more technically correct, database access needs more than one database client connection running in parallel.

Ditto for downstream bitcoind access -- pushpool should probably contact a farm of bitcoind's, not just one.

Interesting stuff. I have considered writing own pool, but experimenting with 50 BTCs per block doesn't sound very tempting. I guess a quick approach would be running multiple instances of pushpool in multiple virtual machines until a more reliable pool shows up.

Good idea, I didn't know about testnet. That will come in handy, thanks.

Good quastion i do not know hope somebody can tell this do Ill believe the bandwidth should be pretty high

ah, netcom.   that was first commercial internet provider in DFW area outside of the crap hourly stuff like compuserve....    but back then in 1990 or so, you could still freeload off of 1/2 the universities anyway, or at least telnet out from their login interface (have some friend at a uni give you a shell acct or something).   there were a few exceptions,  like at the superconducting supercollider, you could log on directly to their VAX (sscvx1.ssc.gov) with a guest account

for the question

did you change your pushpool "requests to us" to the same port that bitcoin uses?  if so, you shouldnt have. 

I remember netcom, I actually was living in Dallas at the time working in the video game industry.. but yea thats not what im using

I was able to figure it out because in the server.json their are two binding addresses for JSON-RPC.. if you leave them both in you get that issue, it wasnt really clear to me at first and nothing in the docs talk about it from what I could read

where do i report a bug in pushpool?

all 3 pools for ELC are having weird reporting bugs. shares are being paid out but reporting is off. no blocks since a difficulty retarget have been reported, and yet there are blocks being found according to elacoind debug log.

stats output:

Paid Shares [All submitted shares from previous rounds which are already accounted and paid for.]
Your Valid: 214843
Invalid: 137

Unpaid Shares [Total of all shares yet to be paid since the last confirmed block until now.]
Your Valid: 1281
Pool Valid: 20961

Current Round Shares [Total shares since last found block]
Pool Valid: -129270

Estimated Per Payout
0.05989123 ELC

Account Balance
2.4865372800000016 ELC

seems wrong. And this affects all 3 pools. The coin has a different retargetting interval to fix a botcher release sky high difficulty (like all altcoins yay!), but now we have this new issue.

ideas?

Hi Jeff,

I read the thread, and saw mention that pushpool doesn't work for pools beyond several hundred GH. This is related to open connections.
Stratum was mentioned as a solution. I believe that stratum is pretty much the only way people connect anymore.
One developer mentioned leaving Pushpool, but several highly respected pool owners have been quite active in this thread.
The last comments on original dev and Jine's fork are both at 8 months ago.

1. Is the limitation still there?
2. Is anyone still actively developing pushpool? If not, I'd love recommendations.


TL;DR: Is pushpool currently dead? If so, what do you recommend?

pushpool cannot keep up with the bitcoin main network anymore, so it is not really being maintained.

Pushpool can't be used with bitcoin mining anymore since it doesn't support the stratum mining protocol. For a number of reasons bitcoinxt should not be run on a mining pool.