Can not sync test-net with 8.26.1

[bastiandriessen]

In order to support the DigiByte developers, I would like to run a test-net on one of my servers. The service does not sync.

The settings in digibyte.conf:

testnet=1
server=1
listen=1
daemon=1
maxconnections=150
disablewallet=1

Running on AlmaLinux 9 (RHEL clone). Can you please advise how to configure and connect to test-net? Using the 8.26.1 production binaries. Should I checkout straight from git instead?

Tail output of debug.log:

2025-11-21T06:16:23Z ThreadDandelionShuffle: Next shuffle scheduled in 1406 seconds
2025-11-21T06:22:30Z Potential stale tip detected, will try using extra outbound peer (last tip update: 66195 seconds ago)
2025-11-21T06:33:00Z Potential stale tip detected, will try using extra outbound peer (last tip update: 66825 seconds ago)
2025-11-21T06:39:50Z ThreadDandelionShuffle: Performing shuffle
2025-11-21T06:39:50Z ThreadDandelionShuffle: Next shuffle scheduled in 80 seconds
2025-11-21T06:41:10Z ThreadDandelionShuffle: Performing shuffle
2025-11-21T06:41:10Z ThreadDandelionShuffle: Next shuffle scheduled in 452 seconds
2025-11-21T06:43:30Z Potential stale tip detected, will try using extra outbound peer (last tip update: 67455 seconds ago)
2025-11-21T06:48:43Z ThreadDandelionShuffle: Performing shuffle
2025-11-21T06:48:43Z ThreadDandelionShuffle: Next shuffle scheduled in 7 seconds
2025-11-21T06:48:51Z ThreadDandelionShuffle: Performing shuffle
2025-11-21T06:48:51Z ThreadDandelionShuffle: Next shuffle scheduled in 1191 seconds
2025-11-21T06:54:00Z Potential stale tip detected, will try using extra outbound peer (last tip update: 68085 seconds ago)
2025-11-21T06:59:15Z Flushed fee estimates to fee_estimates.dat.

[beggsdl]

Try adding to your digibyte.conf file the following:
addnode=47.37.11.100:12026

I have a couple testnet nodes running and am mining on testnet. But I think my node that is exposed externally is down right now.

[bastiandriessen]

Thank you. As soon as I add an "addnode" in the config file, the digibyted service fails to start.

I am still connected to testnet with a different server (198.58.124.101) with version 8.22.0 and that one is still producing blocks. I just can not get 8.26.1 connected to (that) testnet.

Last couple of lines from 8.22.0 testnet debug.log:

2025-11-21T10:48:09Z UpdateTip: new best=22175be2bc2c3674d069ebbc521fbe598b9d398eb8128c47a4aa16b1fff80809 height=5053115 version=0x20000002 algo=1 (scrypt) log2_work=52.190584 tx=5375733 date='2025-11-21T10:47:28Z' progress=1.000000 cache=0.4MiB(76txo)
2025-11-21T10:49:21Z UpdateTip: new best=00000003290ba649a1f46d28a9cd95e2a1349406a38409dc6c3074f325e77376 height=5053116 version=0x20000202 algo=0 (sha256d) log2_work=52.190584 tx=5375734 date='2025-11-21T10:49:04Z' progress=1.000000 cache=0.4MiB(77txo)
2025-11-21T10:49:27Z UpdateTip: new best=44f6e6d506fabb29cef387d1a89cea420cf8ce649898748d0a8cb98801085201 height=5053117 version=0x20000002 algo=1 (scrypt) log2_work=52.190584 tx=5375735 date='2025-11-21T10:49:19Z' progress=1.000000 cache=0.4MiB(78txo)
2025-11-21T10:49:33Z UpdateTip: new best=18cb19eac90ceea8e6eee75c269ded3e2b0f6efa8388f4998fca893dc9f2dfef height=5053118 version=0x20000002 algo=1 (scrypt) log2_work=52.190584 tx=5375736 date='2025-11-21T10:49:25Z' progress=1.000000 cache=0.4MiB(79txo)
2025-11-21T10:49:38Z UpdateTip: new best=0000000566989b2f855fc43be63866ca40e18c17363f1ed53bb4fe699a278456 height=5053119 version=0x20000202 algo=0 (sha256d) log2_work=52.190584 tx=5375737 date='2025-11-21T10:49:28Z' progress=1.000000 cache=0.4MiB(80txo)
2025-11-21T10:50:17Z UpdateTip: new best=000000010ab211dad017220783e3f2af8c64b8da2b18105954e924f4ab812662 height=5053120 version=0x20000202 algo=0 (sha256d) log2_work=52.190584 tx=5375738 date='2025-11-21T10:49:32Z' progress=1.000000 cache=0.4MiB(81txo)
2025-11-21T10:51:05Z UpdateTip: new best=000000030105e174d049ed01f5f1f01509b471b820fd2a21f0d1e1397940a5d4 height=5053121 version=0x20000202 algo=0 (sha256d) log2_work=52.190584 tx=5375739 date='2025-11-21T10:50:12Z' progress=1.000000 cache=0.4MiB(82txo)
2025-11-21T10:51:57Z UpdateTip: new best=000000078df8848e466a817ba0f28133f5fd3f822d1d3db9ce17558394158a8c height=5053122 version=0x20000202 algo=0 (sha256d) log2_work=52.190585 tx=5375740 date='2025-11-21T10:51:00Z' progress=1.000000 cache=0.4MiB(83txo)

[beggsdl]

Ok, I am also connected to that peer. I have had a challenge getting 8.26.x running on testnet. I think part of it might be because I have a number of transactions "pending" in my 8.22.x wallet. If I run that wallet on 8.26, those "pending" transactions go into mempool, but the 8.26 node does not continue to sync and one or both of my miners fails to produce any more blocks. The only way for me to get out of that situation and start producing blocks on testnet again is to switch that one node back to 8.22 and to delete the mempool file on all my nodes (two for mining and one on my regular computer receiving all the coinbase TXs). I am the only one mining on testnet.

[bastiandriessen]

ah so you are generating all the blocks :) Yes there seems to be an issue with running test-net with 8.26.1. I also tried copying the block chain files from the working 8.22.0 testnet (198.58.124.101) to the new server where I (want to) run 8.26.1, but after reading the blocks it just stalls.

Perhaps my understanding is wrong how testnet works. Does it only work for a certain release? Does it require anything in the digibyte.conf to bootstrap it? Having said that, 8.22.0 did not require any special settings. It took like 24 hours before it synced, but after that running fine.

[beggsdl]

It should just work once it connects to a peer. Nothing special should have to be done for it to sync.

[saltedlolly]

It sounds like this problem again: #142

[bastiandriessen]

Thanks Olly. That issue was in regards to the speed of the initial sync of testnet. Main issue of this thread is that testnet does not sync at all, even after letting it simmer for a couple of days. It may be related though.

[saltedlolly]

On a lower spec machine (such as a Pi 4) the testnet would never sync before the "fast start" fix was implemented. I believe this was because on testnet some algos had not been mined for a long time (if ever!) and the difficulty had been driven sky high. Sometimes it took minutes to process a single block, which meant with a 15 second block time, it was permanently playing catch up. Does it sync if you leave it for days? Is it making progress? Check the logs and see how long it is talking to work though each block.

The "fast start" code provided by Zulu pool fixed this and made the testnet usable again.

[bastiandriessen]

Running this in a Linux VM with 8 GB of Mem and 8GB of swap space and more than 100 GB disk available, so resource wise that should be fine. The first post in this thread has the last couple of lines of debug.log. Running 8.22.0 testnet on a different Linux server with similar resources and that runs fine. There is definitely something different with 8.26.1.

[saltedlolly]

My money would be on it being connected to that issue. Without the fast start code, testnet was unusable before. I strongly suspect the fix has broken in 8.26.x.

[beggsdl]

Hello Olly, good to see you.
When I get some time, I will have to try again to get 8.26.1 running. I already tried it a couple times and, for some reason, it broke my mining, even though I was still mining (scrypt and sha256) against 8.22 nodes. I'll try to spin up a fresh 8.26 node instead of upgrading 8.22 to see if I get anywhere with that.

The first time I ran 8.26 (upgraded from 8.22), these 15 transactions (1million each for 15million total) that had been stuck in that 8.22 source wallet appeared as Pending in the destination wallet. But they will not complete. And my mining broke, so no new blocks could be created until I took 8.26 out of the picture and jumped through some other hoops.

I hope to be able to look at this sometime this weekend.

[beggsdl]

I started a new v8.26.1 testnet node to sync from the beginning. After running for about 45 minutes, I am seeing this repeat in the log. It looks like it sync'd the first 118 blocks, but is now stuck.

...
2025-11-21T16:40:48Z Block 93: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Block 94: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Block 95: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Block 96: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Block 97: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Block 98: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Block 99: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Block 100: version=0x20000002, algo=1, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Height (Before): 100
2025-11-21T16:40:48Z Use default POW Limit
2025-11-21T16:40:48Z Block 101: version=0x20000602, algo=3, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Height (Before): 101
2025-11-21T16:40:48Z Use default POW Limit
2025-11-21T16:40:48Z Block 102: version=0x20000202, algo=0, nBits=1e0fffff, expected=1e0fffff
2025-11-21T16:40:48Z Height (Before): 102
....
2025-11-21T17:19:42Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:19:52Z Height (Before): 117
2025-11-21T17:19:52Z Use default POW Limit
2025-11-21T17:19:52Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:20:14Z Height (Before): 117
2025-11-21T17:20:14Z Use default POW Limit
2025-11-21T17:20:14Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:20:39Z New outbound-full-relay v1 peer connected: version: 70019, blocks=3983770, peer=10
2025-11-21T17:20:39Z DEBUG: VERACK processing started for peer=10
2025-11-21T17:20:39Z DEBUG: VERACK processing completed successfully for peer=10
2025-11-21T17:20:39Z Sent Dandelion discovery hash to peer=10
2025-11-21T17:20:39Z SendMessages: Processing 1 Dandelion transactions for peer=10 (relay_txs=true)
2025-11-21T17:20:39Z SendMessages: Sending MSG_DANDELION_TX ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff to peer=10
2025-11-21T17:20:39Z ProcessMessage INV: Got dandelion inv: dandeliontx ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff have peer=10
2025-11-21T17:20:39Z ProcessMessage INV: Requesting dandelion inv ffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff from peer=10 (is_inbound=0, is_discovery=1)
2025-11-21T17:20:39Z ProcessGetData: Peer 10 requested Dandelion discovery hash, marking as Dandelion-capable
2025-11-21T17:20:39Z AddDandelionDestination: Peer 10 confirmed to support Dandelion - inbound=0, outbound=1, current destinations=2
2025-11-21T17:20:41Z Height (Before): 117
2025-11-21T17:20:41Z Use default POW Limit
2025-11-21T17:20:41Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:22:11Z Height (Before): 117
2025-11-21T17:22:11Z Use default POW Limit
2025-11-21T17:22:11Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:22:13Z Height (Before): 117
2025-11-21T17:22:13Z Use default POW Limit
2025-11-21T17:22:13Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:23:27Z Potential stale tip detected, will try using extra outbound peer (last tip update: 2566 seconds ago)
2025-11-21T17:23:37Z Height (Before): 117
2025-11-21T17:23:37Z Use default POW Limit
2025-11-21T17:23:37Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:23:57Z Height (Before): 117
2025-11-21T17:23:57Z Use default POW Limit
2025-11-21T17:23:57Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff
2025-11-21T17:24:19Z Height (Before): 117
2025-11-21T17:24:19Z Use default POW Limit
2025-11-21T17:24:19Z Block 118: version=0x20000202, algo=0, nBits=1e0ccccc, expected=1e0fffff

[JaredTate]

We likely have missing testnet configuruation in v8.26. We are going to need to completely reset testnet with v8.26 for DigiDollar development. The next version will be v9.26 with a reset testnet. Hopefuly should be ready for testing in next couple weeks. At that point it would be greeat to get a lot more people helping with testnet to test/ debug digidollar.

[beggsdl]

Jared, please keep me in the loop about testnet since I've been the primary one keeping it alive for the past several years.