Forum Discussion
NimbostratusLTM pair says "in sync" but items missing on one unit
Have a 2 node LTM pair, where the standby unit status says "in sync". Also when I try to "sync group to device" standby, the action is not allowed as it's reported that "This action syncs the most recent configuration from the group to the selected device. The device you have selected already has the most recent configuration.".
Both a pool and a VIP are missing on the standby unit. These items were created several months ago. Other items that were created after are fully synced on the standby unit. The issue was only discovered due to a failover and the missing pool and VIP were no longer accessible.
Has anyone seen this behavior before (BIG-IP 11.6.0)? What is the best corrective action?
12 Replies
Yann_Desmarest_Nacreous
Hello,
Are you using partitions ? Are you syncing your configuration when connected with an account different than admin ? if yes, try using the admin account.
Did you try to sync the configuration between both members with "Overwrite Configuration" checked ?
- Yann_Desmarest_Can you also check the traffic-groups and make sure that your VIP is in a traffic group hold by both device ?
tickermcse76_16Only using the default partition common. When viewing the traffic-group (there is just one, the default) failover objects, the standby unit has 1 more item than the active unit. The extra item is the node that is missing. How would I go about adding this node to the primary traffic-group failover object list? Why would it be missing in the first place?
TJ_VreugdenhilCirrus
Had this same issue... there where two virtual-addresses configured as: traffic-group "/Common/traffic-group-local-only"
I switched them to "traffic-group traffic-group-1" and did a normal sync. And the number of VIP's now match between systems.
Yann_DesmarestHello,
Are you using partitions ? Are you syncing your configuration when connected with an account different than admin ? if yes, try using the admin account.
Did you try to sync the configuration between both members with "Overwrite Configuration" checked ?
- Yann_DesmarestCan you also check the traffic-groups and make sure that your VIP is in a traffic group hold by both device ?
- tickermcse76_16Only using the default partition common. When viewing the traffic-group (there is just one, the default) failover objects, the standby unit has 1 more item than the active unit. The extra item is the node that is missing. How would I go about adding this node to the primary traffic-group failover object list? Why would it be missing in the first place?
- TJ_Vreugdenhil
Had this same issue... there where two virtual-addresses configured as: traffic-group "/Common/traffic-group-local-only"
I switched them to "traffic-group traffic-group-1" and did a normal sync. And the number of VIP's now match between systems.
Kishore_Kumar_2How to Verify the difference first
- 1) Perform a VIMDIFF on both the config files
- 2) Perform a MD5 Checksum on the config files to see if the values match
- 3) Check the Objects in Active and Standby device
Procedure :
- 1) Check if you save once you make changes in the Active device
- 2) Perform a manual sync
- 3) After VIMDIFF comparison, If there is no change - NO issues, If not look at the changes if they are recent - perform a config sync then Perform a VIMDIFF again.
- 4) if MD5Checksum value is same. no issues, if they are different perform a SYNC from active.
- 5) Check the Daemons responsible for Configsync CSSD, MCPD
https://support.f5.com/kb/en-us/solutions/public/13000/900/sol13946.html
Check Peer status - Config sync IP - Mirror IP
Please let me know if you still have not narrowed down the issue.
- xhairs_248539
I wanted to reopen this as I have come across this same issue. The pairs are labelled "in sync" however one device has a different configuration than the other (an extra VS exists). At this time the active unit has the extra VS and the standby does not.
What causes this issue and how do we fix it? How can we trust the rest of our devices are in sync if we're seeing this problem way past the time this change was ever made?
Thanks!
superuser_22978I do have the same issue. Is there a solution?
- superuser_22978
I broke the trust and rebuilt the HA and it worked
