Mist v0.5.2 - OS El Capitan 10.11.3 - x hours since last block

[Mundy35]

After the upgrade from 0.5.1 to 0.5.2 mist does not sync anymore.

I have the latest version of Geth:
./geth version
Geth
Version: 1.3.5
Protocol Versions: [63 62 61]
Network Id: 1
Go Version: go1.5.1
OS: darwin
GOPATH=
GOROOT=/usr/local/Cellar/go/1.5.1/libexec

and Geth is synchronized:
> eth.blockNumber
1171528

Screenshot from the Developer tools:

Any idea?

[alexvandesande]

How long has it been that way? Sometimes it takes about a minute for my wallet to catch up and find the latest block. If it persists more than that, try restarting the wallet.

[Mundy35]

I have this issue since the upgrade from 0.5.1 to 0.5.2. It has never worked with 0.5.2.
I have already restarted the wallet, even deinstalled and reinstalled everything (wallet + geth).
The thing is that geth is synchronised and it receives perfectly the blocks. The issue seems to come from the UI.

[swt83]

Same as issues #386 #362 #310 #292

[hiddentao]

Thought I had the same problem and that it was due to an older version of geth in the code but was wrong.

[johnpitchers]

Reported here 5 days ago. #366

[clund]

Solved issue with database upgrade by deleting the database and letting it download again. After shutdown and restart of Mist the following message appears:

Node type: geth
Network: main
Platform: darwin (Architecure x64)

...o /Users/clund/Library/Ethereum/chaindata
I0321 13:35:51.122331 15065 database.go:71] Alloted 78MB cache to /Users/christianlund/Library/Ethereum/dapp
I0321 13:35:51.142665 15065 backend.go:314] Protocol Versions: [63 62 61], Network Id: 1
I0321 13:35:51.143143 15065 backend.go:362] Blockchain DB Version: 3
I0321 13:35:51.145186 15065 blockchain.go:214] Last header: #770694 [35c0cb45…] TD=4879785175929831087
I0321 13:35:51.145235 15065 blockchain.go:215] Last block: #770694 [35c0cb45…] TD=4879785175929831087
I0321 13:35:51.145254 15065 blockchain.go:216] Fast block: #770694 [35c0cb45…] TD=4879785175929831087
I0321 13:35:51.145664 15065 handler.go:92] blockchain not empty, fast sync disabled
I0321 13:35:51.153274 15065 cmd.go:114] Starting Geth/v1.3.5/darwin/go1.5.1
I0321 13:35:51.153400 15065 server.go:311] Starting Server
Fatal: Error starting Ethereum: listen udp :30303: bind: address already in use (possibly another instance of geth is using the same port)

Seems like there's very little Zen with 5.2 on Mac.

[frozeman]

You're already running geth, maybe mist couldn't shut it down properly. Please start the "activity" app and quit geth manually and restart

[frozeman]

@Mundy35 the issue youre having, is that the Mist interface can't connect to the node. This could be because the IPC path is incorrect, but than it wouldn't have started at all. Or the IPC connection was killed from the node side (@bas-vk ?). Please restart the wallet and try a few times.

Everytime you see a IPC connection error, than the connection to the node is broken.

[Mundy35]

@frozeman I have tried several times to restart the wallet but this did not solve this issue.

[clund]

Please start the "activity" app and quit geth manually and restart.

I assume you mean Activity Monitor. No Ethereum processes are active. Also tried:

ps aux | grep -i 'geth'
ps aux | grep -i 'eth'

No running processes matched.

It seems like Ethereum sometimes fails to close down properly and leaves the listener open. Any idea of how to shut it down manually?

UPDATE:

geth --ipcpath /Users/clund/Library/Ethereum/geth.ipc
...
I0321 22:12:29.224855 18554 cmd.go:125] Starting Geth/v1.1.2/darwin/go1.5
I0321 22:12:29.225750 18554 server.go:311] Starting Server
Fatal: Error starting Ethereum: listen udp :30303: bind: address already in use

Tried:

lsof -i :30303

and

nc localhost 30303

Both returned nothing.

UPDATE 2:

netstat | grep 30303

udp46 110267 0 *.30303 *.*

Will try to unbind it.

UPDATE 3:

The netstat reveals that a socket is open on 30303 and data is queued for the application associated with it. However, there is no process linked to the socket. This might indicate that Mist does not properly close all sockets on exit.

Active kernel control sockets
Proto Recv-Q Send-Q   unit     id name
udp46 110267      0  *.30303                *.* 

No luck so far in manually closing a socket on MacOSX without a PID.

[swt83]

I can confirm that simply rebooting your computer solves this problem.

[clund]

Indeed rebooting does the trick, but I would hope that a better solution is introduced in later builds.

[Mundy35]

Rebooting my computer with having only Mist running on it solved this problem for me also.
Thank you all.

[lock]

This thread has been automatically locked because it has not had recent activity. Please open a new issue for related bugs and link to relevant comments in this thread.