hi guys how can i make the dude run a backup using command line (to automate this)

There is almost no any available settings for Dude server from RouterOS console.
When Dude running on the Windows PC, backup is available only from GUI.

bummmmmerrrr i wish i couldve automated backups

In windows... The folder Dude\data contains your data create a scheduled task that makes a copy of this folder somewhere else through a batch job. Maybe just copy the whole Dude folder somewhere else just to be safe? I have not tried this so YMMV...

HTH!

wow nice reply.

I am going to try this my self, let you know how it goes...

yeah it works... I would only run this style backup once a week instead of daily since if you run it daily and you break something bad you might not know until it is too late and the corruption is saved in your backup. The whole Dude directory is like 100M so you could easily script it to do multiple different directories or a batch job for each day of the week so you can recover from any point in time ... like 1 week 1 month, you know real backups

c:
cd "\Program Files"
xcopy /Y /S dude dudebackup

excellent and doing a restore from this backup worked out good for you?

Heck no I am not testing that hehe Unless I break something. I suppose I could run it on another computer... Just copy the directory but alas I have no interest in that either.

I have a fairly complicated setup all built by hand.. no auto discovery and do not want to play with the working system unless I am forced to.

Keep in mind if your going to try to copy back you need to stop the service and make sure the exe is not running on the server also.

tested and works perfectly.

Funny I just ran into this thread. rc3 just committed unrecoverable error. Unable to restart instance. MMC 'Stop' service produced a script error in IE! Unable to stop instance...Restart? What do you think?...

BTW, the event viewer suddenly has started reporting ~680 occurrances of a dude restart.
Reporting this error repeatedly, yes over and over again...WHAT! TF....

So, as some were asking, running dude.exe from a copy of the subdirectory of a former instance will certainly work, as has occurred at times to serve as production. Just be careful you remember who is accessing it, and what data they are viewing. It and they could easily get confused, maybe start re-starting itself incessantly....I digress. Also make sure you deactivate 'run as service' for the original instance through the server settings at login. If you dont, I cannot predict if you will be able to unscramble your data, so be careful. Yes, running it on a remote machine, maybe even not networked would be pretty bulletproof for non-corruption of production data. And, just make sure there are no dude.exe procs running. If so, they are up to no good. End them...

What gives Normis? I very rarely lose data, but this was fatal

The only crash report available is the event viewer message. It seems to be happening on a regular basis, about every 10-11min. It must be one object that is corrupt, even though I have rolled back to previous backup before it was experiencing trouble.

Event Viewer log entry:
The The Dude Server service terminated unexpectedly. It has done this 81 time(s). The following corrective action will be taken in 60000 milliseconds: Restart the service.

This message varies, but always with 'It has done this 'too many times' ' !!!

Will send a message to support, but this is killing me, and corrupting my bandwidth history again...

Hey Znet, Damn that sounds bad...

I thought of the trouble from running Dude from the backup/non-default directory while having the other directory still in tact and being the install directory could cause serious badness, is this what happened with yours?

I think people should never run Dude from their backup directory. Just copy files around while the dude is shut down if trying to do a restore from a backup.

Anyhow I hope I didn't contribute to your trouble and that you find a quick solution.

SD

No, I have not lost data by running from the backup subdirectory. However, depending on how your setup is configured, if you accidently boot the machine and forget that the auto-start instance of the Dude hasnt been disabled, you could run into problems starting the backup instance.

I stripped all duplicate objects (on different maps), and pared it down to minimum, checking every credential possible, and it still is misbehaving. Will try it on a totally remote and clean server and see if it chokes. I also will have to try Dude on RouterOS. I have it on a 1U server, and maybe that isnt broken...

OTW, I am forced to recreate. I smell a null string in the data store. Not likely to correct that problem. Restore methodology must be improved to incorporate importing single maps, or other groupings of devices. That way I could have a clean copy of all the objects, that I know are stable, and only restore known good template objects. This can be accomplished by a shaky manipulation of XML data, hopefully avoiding the corrupt data...

Thats what I think. I think Normis should tell us when the 'Selective Import' feature will be implemented...Thats what I think...