12-11-2017 10:37 PM
Hello Folks,
I found a strange behavior of public API for log extraction. Previously, we created our pipeline to handle and extract the error message by using the index ‘0’ of entities. But in recent past, we noticed that the relevant error message is coming on the last index of the REST call to the API. It’s like it worked as queue for collecting the errors before and now it is working as stacks.
Because of this observation and a possible uncertainty, we had to come with a logic to extract error message based on least timestamp in the entities array. Can someone confirm such observation and as well state that it will work as queue or stack or my logic is undeniable? (I, Robot movie dialogue of WIKI 😀 )
Thanks,
Suvigya.
12-17-2017 09:09 PM
This is in regards to querying the logging API to extract ‘ERROR’ for a specific runid upon failure in child pipeline.
When we query we get an array with the error results and per our observation the element index of the array are increased as per the root error’s effect on other components. Earlier the ‘0’ index gave the first (root) place of error. Now its the last element. Any pointer on this?
Any help will be appreciated.