Last Modified: Oct 07, 2026
Affected Product(s):
BIG-IP LTM
Known Affected Versions:
17.1.3.5, 17.5.1.9, 21.1.0.2
Opened: Oct 06, 2026 Severity: 4-Minor
On a multi-blade VELOS tenant, running istats remove, istats reset, or tmsh reset-stats ltm virtual <vs> from the primary blade might not clear the cluster-wide iStats counter value The commands clear the primary blade’s per-TMM values, but secondary blades can retain non-zero values. As a result, istats dump and tmsh show ltm virtual continue to report a non-zero aggregate value. Merged iStats values can also differ between blades
Affected iStats counters retain stale values after a reset or removal, causing cluster-wide statistics to remain non-zero and potentially differ between blades
- A BIG-IP tenant runs across multiple VELOS blades - iStats counters contain values on multiple blades - iStats synchronization to one or more secondary blades is disrupted The issue has been observed on an eight-blade customer deployment. A similar issue was reproduced in a two-blade lab system by repeatedly restarting merged on the secondary blade while modifying iStats on the primary blade. The cause of the customer-observed synchronization failure remains under investigation Other multi-blade platforms, including VIPRION and vCMP deployments, might also be affected, but this has not been confirmed
Restart csyncd on each affected secondary blade: bigstart restart csyncd
None