Forum Discussion

Scooter_42218's avatar
Scooter_42218
Icon for Nimbostratus rankNimbostratus
Oct 31, 2011

Action on service down + Persistence

My company has run into stability issues with a specific application during a server outage. The environment is overbuilt by 10x (planned rollout later this year) and they shouldn't have problems when a single server goes down.

 

 

We found that the action on service down is set to 'none' (yay defaults) and we have plans to move this pool to 'reject'.

 

 

We're trying to determine if this is a change we should roll to a more broad audience, and we are trying to figure out what happens to persistence records when the pool is set to 'none'.

 

 

thanks!
  • With action "none" if the node is going down the client will be stuck to that node even if it´s going down so usually "reselect" is the way to go.

     

    with clients and the option reselect clients will be directed to new node then it depends of your applications if that is OK. Worst case is usually that the client is "thrown out" and needs to log in again if persistence is mandatory.

     

     

     

    Beinhard

     

  • Thanks to the US choosing a totally absurd date format, I'm not sure if this is a really old post, was it 01 November or 11 January?

     

     

    Anyway this may help... There's a good techtip on the behavior of this setting at https://devcentral.f5.com/Tutorials/TechTips/tabid/63/articleType/ArticleView/articleId/179/LTM-Action-on-Service-Down.aspx

     

     

    None (default)

     

    LTM will continue to send data on established connections as long as client is sending and server is responding. Connection management / recovery / cleanup is via standard TCP mechanics for both clientside and serverside flows.

     

     

    Use "None" if you don't want LTM to intervene in managing either side of the connection. Useful if your servers may not be accepting new connections, but should be allowed to continue servicing existing connections when marked DOWN. Also supports custom monitoring designed to support connection bleeding and other non-standard state management schemes.

     

     

    Matt