Community discussions

 
Ehman
Member
Member
Topic Author
Posts: 363
Joined: Mon Nov 15, 2010 10:49 pm

a working failover config needed

Wed Oct 10, 2012 5:36 am

Hi folks

I've testing out 20 fail over scripts tonight, and none seems to work really

I've got a routerboard with 2 WAN connectionsn and I just want WAN1 to be the main, and WAN2 to be the slave and the trigger must be to ping 8.8.8.8 from only WAN1, and if WIN1 is messedup, it must turn WAN2 on for main



I were unable to find anything work


at this moment I've got
WAN1 = distance 1
WAN2 = distange 2

but when it hits WAN2, your screwed, it wont jump back when WAN1 goes on or gets available
 
gsloop
Member Candidate
Member Candidate
Posts: 213
Joined: Wed Jan 04, 2012 11:34 pm
Contact:

Re: a working failover config needed

Wed Oct 10, 2012 10:02 pm

I wrote a script like this:
http://forum.mikrotik.com/viewtopic.php?f=9&t=60247

Have you tried that?
I'm glad to help you get it working if you have issues.

-Greg
- If I helped you solve your problem ... Karma is an appropriate gift! :) -
 
User avatar
tomaskir
Trainer
Trainer
Posts: 1120
Joined: Sat Sep 24, 2011 2:32 pm
Location: Slovakia

Re: a working failover config needed

Wed Oct 10, 2012 11:22 pm

I have a script you can try on the wiki: http://wiki.mikrotik.com/wiki/Failover_Scripting

You should be able to configure it quite easily, let me know if you have any problems.
Unimus - configuration management, automation and backup solution
Mass Config Push, network-wide RouterOS upgrades, and more!
 
newranman
just joined
Posts: 17
Joined: Mon Jan 10, 2011 10:53 pm

Re: a working failover config needed

Thu Oct 11, 2012 8:36 am

 
gsloop
Member Candidate
Member Candidate
Posts: 213
Joined: Wed Jan 04, 2012 11:34 pm
Contact:

Re: a working failover config needed

Thu Oct 11, 2012 9:58 am

@newranman
No scripting needed
True enough, though there are, IMO, very significant differences between the two approaches. The script version is lots more granular. You can set levels at which the pipe will be considered provisionally down, even if some ping replies come back. You can even do some rough RTT measurements and decide the connection is too bad to consider up.

Email notification when the connection is down and when it's up is another.

So, while they may work to the same basic point, they don't approach the issue from the same perspective or solve it in very similar ways.

However, doing so without scripting is perhaps easier.

[Just like grep and find are a lot different than a windows search. They both find stuff, but the flexibility of grep|find is a lot higher, while also being substantially more complex in use.]

I won't say my way is the best, or only way. But it was the best for my application and I thought I'd share so someone else could benefit from it.

-Greg
- If I helped you solve your problem ... Karma is an appropriate gift! :) -

Who is online

Users browsing this forum: MSN [Bot] and 100 guests