Wallet (1.2.9) / Node gets stuck all the time (shown in networkstatus but it seems to be correct), Wallet 1.3.0: good ?

  • Hi,

    i am just running a node with wallet 1.2.9. The wallet gets always stuck then i resync and after some time it gets stuck again and then resync and gets stuck and .....

    Its shown under the network status but the wallet itsself in the terminal (Linux) shows no errors or hints. When there wold be no network status (block explorer) i would think everything is working fine.

    So my question. When my node is shown stuck in network status and there is no message in my wallet terminal could it be that the information of the network status is wrong and the node is running fine ?

    What about the newest maria db walletl. Is it better than 1.2.9 , in the installation information there stand something with a password . Why is there a password (not the password for your wallet account) ?

    When i set sharemyAdress to false then then Wallet runs stable (but no node) and is alway on the right block.

    Somebody some answers ? Are you running the 1.3.0 Wallet ?

  • I'm running 1.2.9 as well, best solution I've found is to have a cron job kill and relaunch the wallet every hour, that way it usually clears any stuck moments on it's own. If it's running as a service you can just call the service to restart.

    You are out of sync for the ~3-5 minutes it takes to resync, but it keeps it going.

    Hopefully someone can make a wallet that knows when it's stuck and can do what it needs to do on its own. From what I've seen even the 1.3.0 wallets get stuck.

  • @BurnMe hm, I rarely had an unstable/disconnected wallet (aside from the July spam attack). Currently running 1.2.9 since release day.

    How many peers do you connect ? I use the following in my nxt.properties:


    I ran a named (full) node for more than 2 years, but since July there are so many clients that resync continously that my outbound traffic was in the Terabytes (>5Mbit/s). Being a "normal" client w/50 connections needs some 6 GB/month (25 Kbit/s). I run 2 wallets on different networks, both have ~1/3 not-up-to-date peers.

    I see the memory consumption at ~9 GB RAM for both (>>20d uptime), but as I have plenty I'm starting with 8min - 16max:

    java -Xms8G -Xmx16G -d64 -cp burst.jar:lib/*:conf nxt.Nxt

    What is the mem usage when your wallet hangs ? OpenJDK-1.8 ?

  • Hi,

    thanks for all answers. The max number of connected public peers is 20.

    The memory consumption was at 2 GB. But all in all (running others programs) with Linux i have to use the swap file. But this should be no problem because nothing is running out of memory.

    I'm using open jdk 1.8, what about 1.9 ? Is 1.9 better for running a node ?

  • @BurnMe You may be experiencing what I was in this thread.

    Set your logging detail to fine in the conf directory and see if you see something similar to what I posted in your logs. My server was getting stuck every 24 hours or so.

Log in to reply

Looks like your connection to Burst - Efficient HDD Mining was lost, please wait while we try to reconnect.