Quote from: clintar on January 09, 2015, 12:23:36 AM

Quote from: mbk on January 08, 2015, 09:35:36 PM

My pool bbr.mbkpool.info had problems with payments lately. It was due to wallet getting out of sync like many times before. All miner payments for last few days have been sent. I rescheduled the wallet restart script to prevent such problems in the future.
Just noticed - cncoin.farm has 97% of total hash rate. It's a reliable pool but it doesn't look good for the coin. Please, use solo miner or some other pool if you can.

Hey, here's a git diff that fixes that if it's that the wallet seems to stop checking for updates.

Code:

diff --git a/src/wallet/wallet2.cpp b/src/wallet/wallet2.cpp
index a525a86..c8d7ee4 100644
--- a/src/wallet/wallet2.cpp
+++ b/src/wallet/wallet2.cpp
@@ -436,10 +436,11 @@ void wallet2::refresh(size_t & blocks_fetched, bool& received_money)
   catch (const std::exception&)
   {
   blocks_fetched += added_blocks;
- if(try_count < 3)
+ if(try_count < 3000)
   {
   LOG_PRINT_L1("Another try pull_blocks (try_count=" << try_count << ")...");
   ++try_count;
+ sleep(60);
   }
   else
   {

I believe that came from Onishin in #monero-pools when he posted it there about monero's wallet doing that a while ago. Haven't had to restart because of this issue since doing this. Now not being able to make a payment because of dust and transaction being too large bites me all the time.

This is not perfect solution, since wallet2 class used in differen components, and this approach could broke or frees GUI for example.
Just commited more correct solution that should work for pools as well.
(The real problem was cz throwing exception that killed RPC server thread)

Thanks to Clintar to point to this problem

Zoidberg