Btw, if you notice that the accelerated DHT client consumes too much resources, consider trying out Optimistic Provide. You can read more in the config docs.
What happened to ReprovideSweep
? I thought that was the best possible solution to the problem. Is that still being worked on?
I believe itās actively being worked on, but not yet available: Amino (the Public IPFS DHT) is getting a facelift | IPFS Blog & News
The first test with AccelerateD was not successful. The machine restarts every few hours at night. Now I have scaled it to 4 cpu.
While looking through the logs, I found this:
ERROR fullrtdht fullrt/dht.go:450 key not found in map
Unfortunately, I cannot reproduce this error. It showed up two days ago.
stats provide output:
{
TotalProvides: 4,
LastReprovideBatchSize: 0,
AvgProvideDuration: 1137900918,
LastReprovideDuration: 0,
}
/api/v0/bitswap/reprovide
- returns timeout. I guess this is not a valid response.
Strange. Iām not sure why this is happening.
Can you try to run the following commands and share the output:
Get a count of blocks in your repo
$ ipfs repo ls | wc -l
Verify all blocks in your repo (to make sure their not corrupt)
$ ipfs repo verify
...api/v0/repo/ls" | wc -l :
282672
.../api/v0/repo/verify" | grep -v '"Msg":""' :
{"Msg":"verify complete, all blocks validated.","Progress":0}
Iāll try āoptimistic provideā today. Unfortunately, aceleratedht and 4 cpu the problem persists. Node restarted after a much longer time.
@michalgrzegorczyk Any update on how it went with optimistic provide?
In terms of resource consumption, it has improved significantly. The service does not restart and has constant RAM/CPU usage
Unfortunately, the files are still inaccessible through public gateways.
stats/provide
output:
{
TotalProvides: 605377,
LastReprovideBatchSize: 305774,
AvgProvideDuration: 599048403,
LastReprovideDuration: 216694024809842,
}