Unexpected panic in ES bulk After
I use alias to handle index version changes, and somehow I assigned 2 indices to 1 alias, but I forgot to define the write index, I got the following panic when I update indices
panic: runtime error: invalid memory address or nil pointer dereference
When I dig deeper I found that current logic in ES bulk After can cause panic if the passed in bulk response is nil.
One of the cases is elastic: Error 400 (Bad Request): no write index is defined for alias...
, which I meet :-).
We should check nil before getting something from bulk response, and check errors first is always a good idea.
Edited by Xing Xin