Hello,
I upgraded to 2.2.1 last night. My backups are set hourly, and I am seeing a few devices run a backup where the entire configuration is marked as deleted and replaced by "/export compact". Then on the next hour, the proper config returns.
So far this has happened a couple times on two Audiences and one hEX. They are running ROS 6.49.2
[Fixed in 2.2.2] 2.2.1 Mikrotik Backups Are Being Replaced With One Line
Hi,
Sorry you are running into this - we would like to hotfix this ASAP. Could you please enable logging (Debug Log & Device output logging in Zones > Your_Zone > Debug Mode) and send us logs from a backup run that retrieves the backup incorrectly?
Please create a Support Ticket on the Portal with the logs attached and we will prioritize fixing this ASAP.
Sorry you are running into this - we would like to hotfix this ASAP. Could you please enable logging (Debug Log & Device output logging in Zones > Your_Zone > Debug Mode) and send us logs from a backup run that retrieves the backup incorrectly?
Please create a Support Ticket on the Portal with the logs attached and we will prioritize fixing this ASAP.
Hi. I have 2.2.2 and still getting the same issue. Upgraded yesterday.
Example.
Backup 1
add action=accept chain=IPv4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=blackhole
changed to
add action=accept chain=IPV4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=\
blackhole
Next backup
add action=accept chain=IPV4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=\
blackhole
changed to
add action=accept chain=IPv4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=blackhole
Example.
Backup 1
add action=accept chain=IPv4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=blackhole
changed to
add action=accept chain=IPV4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=\
blackhole
Next backup
add action=accept chain=IPV4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=\
blackhole
changed to
add action=accept chain=IPv4_IN prefix=0.0.0.0/0 prefix-length=32 set-type=blackhole
-
- Posts: 198
- Joined: Thu Aug 05, 2021 6:35 pm
This is not the same issue as the one described in the topic. This one was related to particular devices which had a problem with echoing a command (displaying the literal command) in the CLI, which took longer than one of our built-in timeouts causing the command itself being considered backup data.
We believe this was an isolated issue for Audience line as we later found out that the particular affected device from hEX line had some unrelated performance issues which would imitate the requirements for this issue to pop up.
As for the issue you describe, we are aware of this issue and we were unable to replicate the issue on our end with our devices (it is very inconsistent), but we are working on our own solution to handle this to handle line breaks regardless of line width (we have been setting terminal width to 200 characters when establishing sessions with MikroTiks).
We believe this was an isolated issue for Audience line as we later found out that the particular affected device from hEX line had some unrelated performance issues which would imitate the requirements for this issue to pop up.
As for the issue you describe, we are aware of this issue and we were unable to replicate the issue on our end with our devices (it is very inconsistent), but we are working on our own solution to handle this to handle line breaks regardless of line width (we have been setting terminal width to 200 characters when establishing sessions with MikroTiks).