Bug #6991

[Keep] multiple response.Writeheader calls

Added by Ward Vandewege about 4 years ago. Updated about 4 years ago.

Status:
New
Priority:
Normal
Assigned To:
-
Category:
Keep
Target version:
Start date:
Due date:
% Done:

0%

Estimated time:
Story points:
-

Description

I'm seeing a lot of this in the logs of production keep servers:

2015-08-14_18:41:17.23735 2015/08/14 18:41:17 [10.XX.XX.XX:55083] GET status.json 0.000731s 200 3981 "OK" 
2015-08-14_18:41:17.27110 2015/08/14 18:41:17 http: multiple response.WriteHeader calls
2015-08-14_18:41:17.27112 2015/08/14 18:41:17 [10.XX.XX.XX:55083] GET index 0.033062s 500 135237 "write tcp 10.XX.XX.XX:55083: connection reset by peer" 

Note that the index call also seems to return a 500.

This appears to be related to data manager: when I kept seeing the above in the logs, data manager kept asking this keep server for an index, and then abort because it couldn't reach another keep server, and then restart, rinse and repeat.

History

#1 Updated by Ward Vandewege about 4 years ago

  • Description updated (diff)

#2 Updated by Ward Vandewege about 4 years ago

  • Description updated (diff)

#3 Updated by Tom Clegg about 4 years ago

This message

http: multiple response.WriteHeader calls

is just telling us that the 500 in the log won't match up with the status code seen by the client for this request (likely 200).

(I'd say we should print something a bit more helpful, like "warning: logging status 500 even though status 200 was already sent to client".)

#4 Updated by Brett Smith about 4 years ago

  • Category set to Keep
  • Target version changed from Bug Triage to Arvados Future Sprints

Also available in: Atom PDF