Tips on staying synced to the Chia network

[bsd43]

I have 4 computers and 1 pi4 4g, recently I move full node a lot among machines for various reasons, mostly kept synced, here is the things I will do for next smooth move:

1. understand router

   this is the most important part from my experience, basically need choose between port forward and upnp, because I have multiple machines, I disable upnp and manually update port forward every times, my router is ddwrt and there’re a lot options to choose, for example one option is to clear port forward before enable upnp, and I struggled a lot to sync before finally understand what this option do.

2. keep a list of good nodes

   • chia show -c > goodnood.txt and edit it to keep the tallest node ip
   • choose from website like chia.powerlayout, choose the tallest node
   • chia show -a ip:port

3. keep the physical copy of the blockchian db and wallet db

   • .chia/mainet/db/* and .chia/mainet/wallet/db/*
   • copy them to new full node
   • if don’t have them yet, there’s one posted in chia subreddit

4. Make sure time and timezones are all correct and accurate on the computers involved.

5. if still not working, go to forum,reddit, discord to ask for good working node, if port forward is right and good nodes are added, no reason it would not sync.

good luck everyone and hope this will help a little

[codinghorror]

Thanks for this! Good tips for everyone.

[hellegaard1]

I got curious and changed values in config.yaml. Target_peer_count, Target_outbound_peer_count, max_peers_to_send, etc. Changed all of them to 200+ and now have a ridiculous amount of full node connections in gui. Haven’t had any problems with desync. For anyone willing to try this, I don’t recommend using more than maybe 2-3x the value that is already there, it may become a security risk. I just figured the more nodes connected to me, the more nodes/peers can sync up from me.

[casualChia]

I’ve been curious about how ones connections into the Chia network affect things like staying in sync or ones reward when a proof is found. Has anyone heard of only tools that provide more insight into the network itself or seen any discussions from those in the know?

[forich]

Wouldn’t it be better to have both Upnp (1 per network) and port forwarding (to 1 machine)?

[hellegaard1]

Upnp is basically automatic port forwarding. It listens for ports that are needed and opens them, when the device is done with the port and after a no activity timeout period, it closes the port. Upnp doesn’t always work correctly though. Using either upnp/forwarding/triggering together is redundant, if Upnp isn’t working correctly, another SHOULD kick in.

[raptuspl]

If you still have problems to keep up a synced status, that lovely green colored word…
If you are synced but the moment you start plotting it goes unsynced…

The Error might be not only a difference in Timezones but also in Location of those!

Let’s assume you are living in Warsaw(Poland) and your Windows Timezone Settings are on GMT +1 (Warsaw, Sarajevo…)
Did you check on your Router or even Repeater Settings? They might be also on GMT +1 but…on a different location likewise GMT +1 (Berlin, Viena etc.)

Not only the Timezones have to match, also the Location in those said settings.

IF those Settings are on the same Timezone but not the exact same locations, the synchronisation becomes unstable or rather sporadic.

in my case my router and windows settings matched up, but my repeater (which is from germany imported) was set on a timezone near Berlin.

Also… open up that damn port 8444 in Windows Firewall Settings, even if you have deactivated it.

i Hope this helped you out

[DjDemonD]

I was failing to sync after moving my /db folder contents to a new machine, so I had most of the blocks already. Turned out the clock wasn’t updating on the new machine, an AcePC T6 compute stick, which I am using as my farmer, so I installed NetTime and set it to update frequently. It’s now syncing.

I noticed it as the “peak time” shown in the GUI was the same as the stopped-clock time.

[Cabanas]

Helo,

Im geting this error and i dont have any clue about how to fix it.

Any suggestions?

2021-05-23T01:47:51.020 full_node full_node_server        : ERROR    Exception:  <class 'concurrent.futures._base.CancelledError'>, closing connection {'host': '127.0.0.1', 'port': 8449}. Traceback (most recent call last):
  File "chia\server\server.py", line 531, in api_call
  File "asyncio\tasks.py", line 435, in wait_for
concurrent.futures._base.CancelledError

[ekeloglu]

Linux users can also check below script.

Basic bash script to keep full node synced

That script was really useful for me before 1.1.6 update. Full node connections are far better with 1.1.6 but if someone still having connections issues, can use my below script. It is very basic script and still open to improvements. Basically check every x second that we have enough full node connections. if we have not, it automatically tries to connect chia offical nodes to improve connection state. please save it with any name with script file. like xxx.sh then chmod 777 xxx.sh ./xxx.s…

[mungospo]

waaay to late for a reply perhaps, but is it safe to go w/ nodes from this list? couldn’t they be trying to push through their version of the blockchain?

[MisterSavage]

I would be happy if my wallet would sync, but after a clean install, then a wallet delete/install, it’s still stuck. A rocky start to this journey.

[Moonub]

I have been suffering so hard since changing. The CPU in my rig.

I had a Ryzen 5 3600 in my harvester. I decided to swap it out to use elsewhere and put a less powerful and I thought more appropriate Athlon 3000G in my rig.

The Athlon based machine just doesn’t seem able to stay synced any more. It is losing the sync every 24 or 36 hours.

I can run a full node sync on a machine sitting next to it, using a 5950x , in a few hours, and then copy the DB across. So I keep repairing it.

Is there a fundamental problem with using the Athlon 3000G to maintain sync? I thought it should be plenty powerful.