SRX Services Gateway
Reply
Contributor
caiyu
Posts: 31
Registered: ‎05-10-2011
0
Accepted Solution

global-threshold and global-weight of ip-monitoring

i am confused about these 2 parameter: global-threshold and global-weight. what is the relationship between them?

if my configuration is like this, how these 2 parameters takes effect when remote ip unreachable?

here is my understanding: if 192.168.1.2 unreachable, it will trigger global-threshold 255 - 192.168.1.2 weight 255 = 0. Then failover will happen. but what is global-weight? it seems that it has no relationship with failover.

 

chassis {
    cluster {
        control-link-recovery;
        reth-count 2;
        redundancy-group 0 {
            node 0 priority 200;
            node 1 priority 100;
        }
        redundancy-group 1 {
            node 0 priority 200;
            node 1 priority 100;
            preempt;
            gratuitous-arp-count 4;
            hold-down-interval 90;
  ip-monitoring {
   global-weight 200;
   global-threshold 255;
   family {
   inet {
    192.168.1.2 weight 255;

Recognized Expert
aweck
Posts: 255
Registered: ‎07-24-2009
0

Re: global-threshold and global-weight of ip-monitoring

In the config snipit above, global threshold determines how many monitored IPs (depending on individual weight) need to failover before before the global weight is deducted from the RG's priority.

Juniper Elite Partner
JNCIE-ENT #63, JNCIE-SP #705, JNCIE-SEC #17, JNCIS-FWV, JNCIS-SSL
Contributor
caiyu
Posts: 31
Registered: ‎05-10-2011
0

Re: global-threshold and global-weight of ip-monitoring

Hi Aweck,

 

thanks for your reply.

please correct if this understanding is wrong.

 

When IP: 192.168.1.2 is undetectable, RG1 threshold = global-threshold (255) - weight (255)= 0.  When RG1 threshold = 0, this cause RG1 node0 priority = node 0 priority (200) - global-weight (200) =0.  Since node0 RG1 priority is lower than node1, this cause RG1 cluster to failover from node0 to node1.


aweck wrote:

In the config snipit above, global threshold determines how many monitored IPs (depending on individual weight) need to failover before before the global weight is deducted from the RG's priority.


 

Recognized Expert
aweck
Posts: 255
Registered: ‎07-24-2009
0

Re: global-threshold and global-weight of ip-monitoring

That's it.

Juniper Elite Partner
JNCIE-ENT #63, JNCIE-SP #705, JNCIE-SEC #17, JNCIS-FWV, JNCIS-SSL
Contributor
h3cisco
Posts: 11
Registered: ‎09-20-2011
0

Re: global-threshold and global-weight of ip-monitoring

[ Edited ]

no no ,RG1 priority is not  node 0 priority (200) - global-weight (200) , It is from RG initially threshold 255 - global-weight=55 ,but RG priority is 55 will don't trigger failover , the RG priority must be reaches 0 , then the failover will take place.

So you need change the global-weight to 255 .

 

The attachment is the screenshot from 11.4 CG abouth IP-monitoring.

Hope it could help you .

Copyright© 1999-2013 Juniper Networks, Inc. All rights reserved.