Been using The Dude for about 2 years now on a RB100AHx4 Dude Edition.
Lots of false positives, unbelievable high latency recorded and unstable upgrade of ROS devices.
No crashes.
Decided to switch to a CHR VM running on ESXi 7/Supermicro server.
Moved the db to the CHR and beside having to redo some hosts or else client crashed, everything went OK.
The reported latency is a lot better and I would say less false positives. Mass upgrade of ROS devices is now smooth.
The client on a Windows 10 machine just freezes daily. I can open the interface, I can see the map and devices, but nothing gets updated and no notifications get sent. I have to disable/enable The Dude via Winbox to fix it.
Tried vacuum but no change.
Db size is 3.3 GB.
Any tips on how to fix this please? later edit:
vaccum; did nothing because for some reason RouterOS uses the primary partition as temp, which is 1 GB, thus the command never executes completely.
Managed to get the DB down to ~4 MB with the following:
Code: Select all
sqlite3 dude.db
DELETE FROM outages;
DELETE FROM chart_values_raw;
DELETE FROM chart_values_10min;
DELETE FROM chart_values_2hour;
DELETE FROM chart_values_1day;
pragma integrity_check;
vacuum;
reindex;
.quit
even later edit: seems that reducing the DB size fixed the freezes on CHR.