Cannot get good peers

Based on what Larry stated, the only change in his setup was upgrading from v1.2.9 (while fully synced) to v1.2.10 (during the dust storm - what didn’t help), and last week to v1.2.11. Looks like for some setups those versions just don’t like something that we have hard time identifying.

The last announcement from Chia about the CAT1 upgrade next week states that v1.2.8/9 cannot recover if they fall behind with syncing. Maybe, the same problem is in v1.2.10/11, but less pronounced.

Those zeros in Up/Down xfrs may be interpreted in two ways. Either there is a problem with the network, or access to blockchain db deteriorated to the point that basically nothing is being written / read from it. Again, the only change was the version update what would point to blockchain db code being broken in those versions.

1 Like

It has been suggested in the other thread with basically identical problem to limit the number of peers.

Maybe you can add this line to your config.yaml (I didn’t see that entry in what you provided):

target_peer_count: 20

1 Like

So I read the CATS info and they really want everyone up to 1.2.11. Uggggg…The suggestion of the yaml file to 20 is fine but the problem is the peers I’m getting are just not providing any up or down data for the most part. That is my problem. I’m getting garbage from my peers. No quality.

I was at 717K last night I’m now only at 759K from overnight. The wallet synced back to where it was after I deleteted it so the wallet is moving along, but the DB to be synched is molasses on a cold winter day.

Please read again what I wrote about those peers. On the surface, it appears that those are bad peers that are not providing you with any data. However, the problem is not those peers, but your setup (for so far unidentified reason).

The number of those peers may be overwhelming your db with redundant requests, but is actually not them not cooperating, but rather your node is not accepting / sending any data on time, and thus they are stalling, or dropping off.

Again, that change in number of peers is simple, and you may try either 10 or 20 peers to see whether that makes any difference.

1 Like

I just farmed another block this morning over WiFi and 4G Mobile. I have never had a problem syncing. I have never had a problem upgrading.

The Alt-Coin Pump-And-Dump FUD Patrol will get you running in circles.

Better to isolate the problem and then proceed.

While I don’t need a VPN, it is a quick way to check against ISP network configuration. It does not significantly slow you down, as it is hardware accelerated.

NordVPN is good. BitDefenderVPN is ok.

You stated your networking connections are fast, so of course that’s why I’m mostly interested in configuration.

If you have decent hardware and a clean install, what else is there?

SO one thing of interest…I did a speed test on my disks as something is just not right. My system disk which is a 2GB Enterprise Seagate SATA drive came in at 8kb transfers of 4mb seq write and 10mb seq read and 4kb transfers of 3mb seq write and 5mb seq read.

I ran the test against my D drive which is 10GB SATA Enterprise drive and it gets 182MB/192 MB on 8kb seq read write and 140/133MB on 8KB transfers.

I ran the test against my nvme and it gets 1428/1352 write read on 8kb transfers and 925/725 on read/write of 8kb transfers.

Needless to say, my system disk is much slower than my other HDD which is very odd. I want to re-install windows but I cant due t the format…How do I start brand new out of the box so I can get my os on the nvme? It seems that even though this is not an issue, my DB error logs are saying it is taking 8-11 seconds to complete make sure you DB is on a fast disk…I can’t believe my OS dis is that slow, but it does seem lethargic at times when you start up any program.

My node connected to yours without any problems, and just sent 1MiB to your node (slow, comparing to what it sent to other nodes).

That would support the statement that your node is struggling with doing db updates; therefore, everything is really slow. Again, two things to improve db speed. First is to use a fast media to hold it (I believe you are already on an NVMe). The second is to reduce the number of peers (reduce read/writes load on db).

Try to drop that number of peers to 10, restart Chia, and let’s see what you get.

I ran the tests against all of my other HDD’s and they all have very good throughput. My OS disk to me is an issue.

It is not about your OS drive being slow, but the drive that holds your databases. I though that you moved your dbs to that NVMe.

PM you. :slight_smile: I could not get the link to my db to work given any syntax

Clean / format ssd, install windows from disk.
Reinstall chia / drivers etc.

Sure, no problem. Let’s try it once mre.

  1. Shut down Chia / reboot
  2. copy your dbs to that NVMe
    2.1. create chia directory on that NVMe (assuming x:)
    2.2. create chia/db on that NVMe
    2.3. copy content of %userprofile%/.cha/mainnet/db to that new x:/chia/db
    2.4. create chia/wallet on that NVMe
    2.4. copy content of %userprofile%/.cha/mainnet/wallet to that new x:/chia/wallet
  3. change your config.yaml to point to this new location
  4. once you are done with those changes, start Chia

changes to your config.yaml:

full_node:
  database_path: x:/chia/db/blockchain_v1_CHALLENGE.sqlite
  peer_db_path:  x:/chia/db/peer_table_node.sqlite
  target_peer_count: 10

wallet:
  database_path:     x:/chia/wallet/blockchain_wallet_v1_CHALLENGE_KEY.sqlite
  wallet_peers_path: x:/chia/wallet/wallet_peers.sqlite

it wont allow me due to the current format.

Forget for now about reinstalling your OS. Just too many moving parts.

To restate an earlier point: a “fast connection” is not a sufficient test of networking hardware nor configuration.

I have observed a person fry a fanless cable modem in a hot Summer computer room; to the point where it was still “fast”, but would randomly drop connections. The client argued forever even though I showed him all the tests (and software resets) failing, until finally a new cable modem fixed the problem immediately. (LOL)