OMG. Fulcrum, you suck.
Fulcrum docs say btc indexing needs "slightly more" than ~40GB of space. It's been indexing for DAYS @ ~10 blocks/s. But it crashed overnight because it maxed out the 100GB data partition I allocated for it!
Fine. Set up a bigger partition. Restart. And... ARGGHHH!!!
Again: DAYS of indexing... poof.
And you get to look forward to NEVER hard rebooting your system otherwise you will likely have to redo the whole process. Hope you have a UPS.
Protip: when you are done syncthing make a separate copy of the database so that in case it gets corrupted later you don't have to start from scratch again.
When I set up Fulcrum, I had a similar experience. I tried using the experimental fast mode, and it would crash after days of indexing.
Finally got it to work by NOT enabling fast mode in the config.
That makes sense. My understanding is that fast-sync holds more in memory. So if there's a sudden crash, the current progress has nothing on disk to recover from. I think I only saw about a 20-30% difference with it enabled. Not worth losing more days over.
Yeah, give it a shot without the boost. I think that’s what causes the crashing. At least it was in my case over a year ago.
ElectrumX continues to treat me well, except for the undocumented biannual database compaction issue.
what hardware are you using? my mini pc with 6c ryzen 5500u and 16 gb ram finishes indexing in maybe little more than 12 hrs
It’s a bitch to get going. But once it’s completed the performance is amazing.
Especially if you’re using whirlpool.
Also it was designed for Ressource rich machines…
Not sbc
The laptop is better than a Raspi4 but obv still closer to that end of the spectrum than a more powerful, new desktop or server. Made some settings tweaks, hoping for a better, faster outcome for Round 2.
4 GB raspberry pis are especially prone to crashes. disabling fast-sync helps but i still saw crashes. as somerfeld recommends, during initial sync i did nightly backups of db so i didnt have to start from scratch on crash. in my case, cause of crash was bitcoind connection error; may have sthg to do with communication with ssd. in any event, what i ended up doing was monitor fulcrum log and stop fulcrum for maybe an hour or so whenever i started seeing bitcoind connection errors.
