Anton Chtepine

I upgraded to 1.1.10, encountered the same issue. Then rolled back to 1.1.8, but geth still fails to stop gracefully and times out upon SIGINT

what does this `Downloader queue stats` line actually mean?

I am not sure that `Downloader queue stats` is the actual culprit. It seems more like the result of the problem. when this happens, the node start lagging several blocks...

> Can you do profiling when it lags, and attach the profile log here? Do you mean running the node with `--prpof` ?

@KeefeL I have these detailed logs in the vicinity of the lag: here we have the last block before the lag: ``` ... Feb 09 08:44:15 DS7959 geth[652106]: DEBUG[02-09|08:44:15.362] Snapshot...

happens every 10 minutes

here is another piece, where `Downloader queue message` accompanies a 30sec lag: ``` │Feb 09 10:50:57 DS7959 geth[676510]: INFO [02-09|10:50:57.179] Imported new chain segment blocks=1 txs=149 mgas=26.322 elapsed=73.092ms mgasps=360.118 number=15,103,056...

@KeefeL I have 2 identical (hardware-wise) nodes in the US (1.1.8) and in Germany (1.1.7). German node did not have any isuues up until about a week ago, but now...

@KeefeL any comments on the above?

@KeefeL I am still having this issue. re-syncing from the latest snapshot had no effect