Would like to know the best way to reindex knots on my umbrel node,
And how it’s done,
Also is this a common issue with knots, never seem to have had this issue when I was using Core
It happens both on Core and Knots (they are nearly identical). When not shut down cleanly (such as during a power outage, electrical surge, or a manual power-off of the server), the blockchain data gets corrupted, and when it checks the latest blocks upon re-starting, it sees their checksum is invalid and requires a re-index.
Oddly, I seems like Umbrel did not add a UI for reindexing. I did find on my Umbrel that there is an Advanced tab under the Bitcoin Knots settings page. There I can see a textbox for entering custom bitcoin.conf overrides. It says you can enter comments or key=value lines there.
You should be able to enter either:
# Full reindex (rebuilds the block index from blk*.dat)
reindex=1
or:
# Rebuild just the chainstate (UTXO database)
reindex-chainstate=1
Use only one of these two at a time (depending on what needs reindexing). Once the reindex is completed, be sure to remove the line (or set it to 0) before restarting (otherwise it will reindex again every time it restarts).
Ok, will it matter if I have 2 nodes on the same network when doing the re-index process
No, it shouldn’t be a problem.
Ok, i seen comments that running nodes on the same network could cause conflicts of some type when sinking nodes
Also has there been any changes in using Datum on umbrel, making it easier to setup on umbrel
I’ve never experienced this. I have had dozens of nodes syncing at once before (I currently have two syncing, and they are on the same network as all of my other nodes). I only have one Umbrel, though, so if there is some specific issue with multiple Umbrel nodes syncing, I wouldn’t have seen it.
I haven’t run Datum Gateway on Umbrel myself (I connect my ASICS via Datum Gateway on one of my Start9 servers), so I’m not sure what updates have been happening in that space.
Great that is good to know
Thanks
Trying to reindex and knot continues to stop running, would I have this same issue if I switch to bitcoin core instead. and are these issues on umbrel , I’m not having this issue with start9
You’ve tried both the options, I assume? If it keeps happening, then it is most likely either a hardware problem (the drive going bad would be the most likely, but possible for there to be a problem on the device), a power problem (the PSU going bad and not providing enough current), or a temperature problem (venthilation dusty, fan motor failing, NVMe overheating).
have only tried the chain state option, will try the full reindex option to see what happens , I’m am running other apps on my node related to bitcoin , should I stop those apps , also when I start knots I have peers connecting to it, would that cause an issue,
I will attached my last log
Not able to add txt file , how should I share my log file
Oh, do you mean when you have the option in the config, it doesn’t seem to do anything (i.e. it behaves the same as when it isn’t in there)? If so, I wonder if it is ignoring the config. For the log, maybe just a screenshot of the messages around the time when it stops (or if you see anything in it that looks potentially interesting)
Not sure im understanding your question , im going into the config section in the settings and typing in the reindex chainstate command. And restart knots , it gets to about 39 per cent and stops. I was wondering if I should stop all knots dependency apps, while it is indexing or this may not cause if from stopping
Yes, that is what I was asking. So it does appear to be trying to resync. If it gets to the same point and stops every time, that could indicate a hardware or thermal problem. Stopping everything else while it syncs might be a way to rule out the latter (not necessarily, though, since block verification is pretty resource intensive all on its own).