The process to sync the plutus-pab chain-index for the testnet takes over 3+ days now and will take longer as time goes on. Is it possible to bootstrap the chain-index sync process to reduce the manual sync process? I've tried copying over the chain-index.db (and referencing that file in the chain-index-config.json file) that was already sync'd from another machine, but it appears to be starting over again. Thank you.
2
-
Copying over chain-index.db should work. What exacly are you running on the command line?– Sjoerd VisscherJan 31 at 11:30
-
Thanks for confirming. The command I am running is from the plutus-pab/test-node readme. cabal exec -- plutus-chain-index --config testnet/chain-index-config.json start-index.– lley154Jan 31 at 17:04
Add a comment
|
2 Answers
Chain-index tries to use the existing db file, but if anything goes wrong it just starts over, so it's hard to diagnose. It could be a version difference. Or it could be a silly mistake, but without error it's hard to say.
You could try simple things like running it for a bit on the new machine and copying the file over to the old machine and see if it picks that up.
-
1Thanks for the answer and yes, it is a differen't version of chain-index that I am trying to bootstrap, so most likely that is what the issue is.– lley154Feb 2 at 14:17
It seems, that syncing chain-index.db is faster on a Linux in a Windows Subsystem for Linux (WSL2) than in a standalone Linux. This might be caused of a better virtual memory manager on Windows.
On my Windows 11 Home / WSL2 Ubuntu-20.04 system with 16 GB RAM and AMD Ryzen 5 3500U 2.10 GHz processor and SSD, syncing on 2022-04-20 took about 14 hours.
Kernel version of my WSL2 (uname -r): 5.4.72-microsoft-standard-WSL2 of 2021-06-07, see Kernel Updates.
You can download my chain-index.db file (zipped: 5 GB, synced 100% on 2022-04-20 19h16m UTC, plutus-chain-index 0.1.0.0). These are the logfile and the sha-256 hashes of zipped and original file. Unzipping file with unzip (not gunzip)!
