EXOS Stacking and sys-recovery-level

  • 0
  • 1
  • Question
  • Updated 3 years ago
  • Answered
Over the last time i have some problems that X450-G2 Stacks and mysterious reboots. Through deeper investigation sometime we have some problems with daemons (for example netlogin / policy daemon). Another reboot was happening as we replace a member stack unit.

Currently i suspect the sys-recovery-level which is set to "reset". Is it possible that if some error happens on a stack member that the master unit of a stack is reseting (because of sys-recovery-level) ??

Does it make sense to change sys-recovery-level from reset to none? Any experience which that on recent EXOS Stacks ?

Mainly i use the EXOS 16.1.x version.


recommendations are welcome.
Photo of M.Nees

M.Nees, Embassador

  • 9,498 Points 5k badge 2x thumb

Posted 3 years ago

  • 0
  • 1
Photo of Prashanth KG

Prashanth KG, Employee

  • 5,300 Points 5k badge 2x thumb
Hi Mathias,

Sys-recovery-level should not be a concern here. Because, this configuration would be triggered only for the errors related to the failure of hardware/ any slot failure messages. 

Are you referring to any process crashes when you refer to the netlogin/policy daemon issues? 
It would be interesting to analyze the logs prior to the reboot and understand the cause of the reboots. 
Photo of M.Nees

M.Nees, Embassador

  • 9,498 Points 5k badge 2x thumb
Hi Prashanth,

my concern for this post is to understand the purpose and function of sys-recovery-level. What does this command do espacially on a EXOS Stack (for example X450-G2). in manual only the standard chassis / slot system is metioned.

Does it make sense to change the default setting of sys-recovery-level especially on a stack ?

Regards
Photo of Prashanth KG

Prashanth KG, Employee

  • 5,300 Points 5k badge 2x thumb
Hi Mathias,

Check if the below article helps your query. 

Sys-recovery-level options and their explanation 

It is better to leave this setting to default in a stack as the switch resets the slot upon any HW failure or any critical slot failure messages. 
Note: The explanation in the command reference guide is applicable to a summit stack as well! 

Hope this helps!
(Edited)
Photo of M.Nees

M.Nees, Embassador

  • 9,498 Points 5k badge 2x thumb
Hi Prashanth,
thanks for the answer - now i am able to estimate what is best for my environment (regarding sys-recovery-level).

Regards