-
-
Notifications
You must be signed in to change notification settings - Fork 3.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
monerod sync stuck sometimes. #2795
Comments
If you just say "it's stuck sometimes", it just doesn't help, and isn't going to get fixed. |
I'm currently running 16 monerod instances with |
OK, thank you. I'm fine waiting if I know you're trying to get some info :) |
Just in case, when it gets stuck like that, get a full thread stack trace:
Replace /path/to/monerod with the actual path first. |
@moneromoo-monero OK. Thanks. |
@moneromooo-monero Hi there - I'm facing the same issue: the blockchain on one of my computers was about 40 days behind - when I run the daemon (v11.0.1.0) it syncs 5000 to 10000 blocks then stops syncing although it's not done yet. When I kill the daemon and restart it the syncing process starts normally. Full stack trace (will post level 1 logs if/when the issue happens again):
|
When this happens again, please post the output of those commands: And that trace makes me think you should run with log levels: 1,net.cn:DEBUG |
Something conspicuously absent from this trace is tools::threadpool threads. Did they all die? We're going to need to see more verbose logs... |
@hyc Unfortunately the sync finished on the third attempt (when I put the log level 1 on) so I can't reproduce it any more. If there's a way to bring the blockchain back to a lower height, I can try to resync from the height where I was before the issue to see if it happens again. |
monero-blockchain-import --pop-blocks N vary N as needed |
Here's gdb output
bitmonero.log
sync_info cal
get_info
get_connections
getlastblockheader
|
monero-blockchain-import --pop-blocks 10
I don't know what this command is for. |
Do you have the string "Exception:" in your bitmonero.log ? If so, can you paste them ? |
Yeah there's some exception logs few hours before the sync stop
|
That's exactly the one I was suspecting. It's supposed to be fixed by f2939bd, do you have this one ? |
It's not in 0.11.1.0, which you said you were using. I thought this bug was pretty rare, so I did not include it as I was not sure 100% sure it would not cause side effects. So you can apply it on top of 0.11.1.0 and run that. |
Okay. I updated one of my instance. Will monitor it for few days. |
It stuck again from new build.
|
And you've double checked this is running f2939bd right ? |
Sure. |
Use the patch below, then run the "bad" node again with
|
@moneromooo-monero |
See https://github.com/monero-project/monero-core for this. |
Without any proves I just want to post my impression about it: when synchronizing seems stopping the priority of network connection changes from in to out - means the program is sending much more than receiving. I stopped the demon then and started again, then it's synchronizing again on my own PC. I assume there is a process that is sending data for others to synchronize, but that should be blocked till the own database is synchronized. |
@miningpoolhub why don't you just copy the synchronized databases from one device to another? |
Upload should not have impact on failing to sync. |
yeah as newbie I never can tell much about it. But when I synchronized it happened at the same time. |
@miningpoolhub I too am experiencing the issue: @moneromooo-monero I'm using the latest release version here: https://github.com/monero-project/monero/releases/tag/v0.11.1.0 (did not compile from source) for Ubuntu 16.04 I am going to restart with @miningpoolhub are you building from source using the 0.11.1.0 tag version? or are you using the prebuilt binaries? what OS are you using? can you give more info about your environment please... |
also, fwiw, it seems this guy had the same issue as well @miningpoolhub #2536 |
@razorman8669 I built from source code. Seems exactly same thing is happening like you. |
#2936 fixes the original problem differently. Run half your daemons with that one, and half without, and compare. |
Any luck ? If no reply after some reasonable amount of time, I'll deem this fixed. |
@moneromooo-monero I have had the |
OK, I'll keep this open then, add a comment when you stop testing. |
#3308 also fixes a sync failure. |
i had this buggy behavior on linux with 0.11 and that was fixed with 0.12. now after running one week, sync_info reports only alive peers. |
Is this only an issue with 0.11, or is it still present in 0.12? I'm noticing similar behavior on 0.12, running monerod under Windows Subsystem for Linux for the first time. What I'm noticing is that running monerod under WSL does work (I got the firewall and port forwarding working), but it seems to slowly lag on my machine (not a fast machine, mind you), getting progressively behind the current block count over a number of hours, to the point where, if I don't restart the daemon, it will finally lock up where typing commands into it is unresponsive. I was thinking that this could just be that my machine isn't fast enough to handle the virtualization overhead of WSL, but then I came across this bug report, which sounds very similar. If I restart the daemon, it starts syncing properly again, before starting to lag yet again. |
Sync stuck today at 1565706 again. Someone reporting this issue on reddit. |
AFAIK, all such bugs are now fixed in 0.12.2.0. +resolved |
I submitted same issue before and it was closed.
#2649
It stuck again. I'm running 0.11.1.0 Helium Hydra Point Release 1.
I did not run with --log-level 1, so there's no log data to show currently.
I just want to reopen this case. and check if I'm the only one experiencing the issue.
I'm currently running more than +10 monero server instances and only one stuck.
It runs fine when I just restart it.
The text was updated successfully, but these errors were encountered: