Support Questions

Find answers, ask questions, and share your expertise
Announcements
Celebrating as our community reaches 100,000 members! Thank you!

ERROR- HUE - Results have expired, rerun the query if needed

avatar
Expert Contributor

queries in hue using impala are returning the following error:

 

"results have expired, rerun the query if needed"

 

as shown in the screenshot below:

query_error_hue_with_impala.PNG

what could be causing this error and how to fix this issue?

1 ACCEPTED SOLUTION

avatar
Master Collaborator

Looks like there are different types of AWS Load Balancers, and the one that can handle TCP sessions are either Network Load Balancer (NLB) or Classic Load Balancer. Which one are you using? I believe stickiness also works for NLB, but do validate with your AWS team.

 

The session timeout is two-fold:

  • LB session timeout. These are settings inside of the load balancer, nothing to do with Impala itself. The recommendation is to set this timeout to 12 hours, 6 hours at a minimum for both client and server timeouts.
  • Hue session timeout (idle_session_timeout). This is how long Hue will keep the connection to Impala alive. Default may be 15 minutes. If no queries run in this time amount of time, and then you run a new query, you'll get this error "Results have expired". Hue will need to start a new session. Also if your query takes longer than idle_session_timeout, then you'll definitely need to increase the timeout setting. 1 hour timeout may be appropriate here.

 

Regards,

Alex

 

 

View solution in original post

6 REPLIES 6

avatar
Master Collaborator

Hi @yagoaparecidoti ,

 

If you have Impala daemons that are load-balanced (LB), then there is a configuration that needs to be applied on the LB, related to source IP persistance. There is a Knowledge Base article about this: https://community.cloudera.com/t5/Customer/Load-balancing-between-Hue-and-Impala/ta-p/73364

 

If there is no LB at play here, then has this issue been ongoing or just started to pop up recently? What changed on the cluster recently?

 

 

Regards,

Alex

avatar
Expert Contributor

hi @aakulov 

 

the cluster is on AWS and the AWS LB is being used on ports 21050 of the impala daemon, in which case this LB may be having this problem?

 

sould it be that in this LB we can also change it to "source"?

 

avatar
Master Collaborator

That's right, you'll need to look into how to setup AWS LB to have sticky sessions, which is not enabled by default. Also, please look into tuning the session timeout to a high value (long enough for Impala to finish the query and send the results back to Hue)

 

Reference: https://docs.aws.amazon.com/elasticloadbalancing/latest/application/sticky-sessions.html

 

 

Kind regards,

Alex Akulov

avatar
Expert Contributor

LB AWS sticky sessions settings is for LB HTTP/HTTPS only

 

in this case the LB we are using is TCP on ports 21000 and 21050

 

the session timeout you say in the impala service, right?

 

would be the "idle_query_timeout" and "idle_session_timeout" options, right?

avatar
Master Collaborator

Looks like there are different types of AWS Load Balancers, and the one that can handle TCP sessions are either Network Load Balancer (NLB) or Classic Load Balancer. Which one are you using? I believe stickiness also works for NLB, but do validate with your AWS team.

 

The session timeout is two-fold:

  • LB session timeout. These are settings inside of the load balancer, nothing to do with Impala itself. The recommendation is to set this timeout to 12 hours, 6 hours at a minimum for both client and server timeouts.
  • Hue session timeout (idle_session_timeout). This is how long Hue will keep the connection to Impala alive. Default may be 15 minutes. If no queries run in this time amount of time, and then you run a new query, you'll get this error "Results have expired". Hue will need to start a new session. Also if your query takes longer than idle_session_timeout, then you'll definitely need to increase the timeout setting. 1 hour timeout may be appropriate here.

 

Regards,

Alex

 

 

avatar
Community Manager

@yagoaparecidoti Has the reply helped resolve your issue? If so, please mark the appropriate reply as the solution, as it will make it easier for others to find the answer in the future. Thanks!


Regards,

Diana Torres,
Community Moderator


Was your question answered? Make sure to mark the answer as the accepted solution.
If you find a reply useful, say thanks by clicking on the thumbs up button.
Learn more about the Cloudera Community: