You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Don't force flush state files on nodeos exit. When state files are on a zfs disk with lots of I/O (this a large server with many applications on it), it can take a long time for all the data to flush to disk. In this example, it took 30+ minutes to stop nodeos
Nov 7 13:20:21 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 94% complete...
Nov 7 13:20:22 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 94% complete...
Nov 7 13:20:23 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 95% complete...
Nov 7 13:20:24 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 95% complete...
Nov 7 13:20:25 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 96% complete...
Nov 7 13:20:26 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 97% complete...
Nov 7 13:20:27 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 98% complete...
Nov 7 13:20:28 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 98% complete...
Nov 7 13:20:29 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, 99% complete...
Nov 7 13:20:29 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, flushing buffers...
Nov 7 13:42:29 eos-sfdm41 nodeos[1080]: CHAINBASE: Writing "state" database file, complete.
Nov 7 13:52:33 eos-sfdm41 nodeos[1080]: info 2023-11-07T13:52:33.557 nodeos http_plugin.cpp:479 plugin_shutdown ] exit shutdown
Nov 7 13:52:33 eos-sfdm41 nodeos[1080]: info 2023-11-07T13:52:33.750 nodeos main.cpp:185 main ] nodeos successfully exiting
This is nodeos 3.2.4
The text was updated successfully, but these errors were encountered:
Flushing on exit is important to maintain consistency. The new mapped private mode in 5.0 may shutdown more quickly. I wonder if that would address your pain point?
Don't force flush state files on nodeos exit. When state files are on a zfs disk with lots of I/O (this a large server with many applications on it), it can take a long time for all the data to flush to disk. In this example, it took 30+ minutes to stop nodeos
This is nodeos 3.2.4
The text was updated successfully, but these errors were encountered: