"REORGANIZE: Block is causing a reorganize" causing btcd to crash on testnet #1492
Comments
|
Time heals all wounds the say. Not in this case - still crashing on restart: What could this mean |
|
I am also seeing a similar issue. It does not happen on every reorg, but I have seen it crash my testnet node twice. The stack trace is nearly identical but I can't seem to pinpoint why it is occurring. |
|
Not sure if this will help, but since both of our stack traces include 2 notifications, I think this might do the trick if you want to port it over to btcd. |
|
Thank you for reporting this significant issue! I suspect that line 283 (cf. line 278) in btcd\mempool\estimatefee.go is actually the culprit, potentially due to a negative index. As you can see, the panic occurs before the locking issue. |
|
@cubancorona can you please let me know if you have a fix for the mempool code? Is this something you are actively working on? Also, I would assume it isn't just a negative index. Usually go will output that error explicitly! |
|
Another of my nodes crashed with the same error
In this case I was able to restart btcd and continue syncing but this is really getting unsettling here... Would you suggest a downgrade to mitigate this? |
|
Hi, all, I have been testing a potential remedy for the last few days: |
|
Can we please have an update on this? Another two testnet3 nodes died with |
|
Hi, @kepikoi, thank you kindly for requesting an update on this potentially important issue! What commit or btcd version are you running? If you can duplicate the issue, would you mind helping me test the following commit, containing major modifications, some of which are intentionally speculative? Thank you again for following up! |
|
@zquestz, please forgive me for not responding directly to your earlier request. I agree that, if my analysis is correct, the stack trace was a red herring in some significant mannerisms. I have not been successfully duplicating the issue, so it's hard to speculate! I am hoping some additional testing will help me confirm the issue. Are you able and willing to help me further isolate the potential issue? Itadakimasu! |
|
@cubancorona I have not experienced this issue since I put gcash/bchd#308 into bchd. Since it is hard to reproduce I am not 100% sure it fixes the issue, but I haven't seen it since I put in the fix. =) |
|
@cubancorona currently I'm running f3ec130. I did restart btcd and unlike in #1492 (comment) this time both nodes were able to process I'm willing to try cubancorona@5b03458 on a designated instance and will report the next time a |
|
@cubancorona In the past weeks I had at least one occurrence on my vms but the cubancorona@5b03458 instance ran stable. So maybe it fixed it? |
|
A few hours ago all (2) of my |
|
https://forkmonitor.info/ is having the same issue |
|
This happened on the @BitMEXResearch ForkMonitor node too. I think it happened before too. Running version 0.21.0-beta. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
My testnet instance is crashing repeatedly trying to sync block 1,608,354 with following log
CHAN: REORGANIZE: Block 00000000d9356478290d6b51b176e0c7b320e4de22090f5c1f0b4b9ff80cbcae is causing a reorganize.. I'm running 0.20.1 with --addrindex enabled.Here an excerpt from the logs:
It seems to be an issue with only one of my identically configured nodes (Ubuntu 18.04.3 LTS)
The text was updated successfully, but these errors were encountered: