[
https://issues.apache.org/jira/browse/DRILL-3939?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14959436#comment-14959436
]
Deneche A. Hakim commented on DRILL-3939:
-----------------------------------------
If you look at the "Pool" field, you'll notice that it's defined as a map first
then as an array of maps. I don't know if Drill should handle this properly but
it seems for it treats this as an unsupported schema change.
Changing the first "Pool" field to the following fixes the problem:
{noformat}
"Pool": [{
"myPool": "PoolName"
}]
{noformat}
> Drill failed to pars valid JSON object
> --------------------------------------
>
> Key: DRILL-3939
> URL: https://issues.apache.org/jira/browse/DRILL-3939
> Project: Apache Drill
> Issue Type: Bug
> Components: Functions - Drill
> Affects Versions: 1.0.0, 1.1.0, 1.2.0
> Environment: Redhat Linux 6.7 Java 1.7 , 1.8
> Reporter: Jaroslaw Sosnicki
>
> The following valid JSON object queried from DRILL using various clients:
> --- t.json start---
> {
> "l1": {
> "f1": "text1",
> "f2": {
> "command": "list",
> "StorageArray": [
> {
> "array1": "Array1",
> "Pool": {
> "myPool": "PoolName"
> }
> },
> {
> "array2": "Arrays2",
> "Pool": [
> {
> "myPool": "PoolName1"
> },
> {
> "myPool": "PoolName2"
> }
> ]
> }
> ]
> }
> }
> }
> --- t.json end ---
> Generates the following error:
> ----
> ERROR [HY000] [MapR][Drill] (1040) Drill failed to execute the query: SELECT
> * FROM `dfs`.`hdvm`.`./t.json` LIMIT 100
> [30027]Query execution error. Details:[
> DATA_READ ERROR: You tried to write a Map type when you are using a
> ValueWriter of type SingleMapWriter.
> File /mapr/demo.mapr.com/data/hcs/hdvm/t.json
> Record 1
> Line 16
> Column 39
> Field Pool
> Line 16
> Column 39
> Field Pool
> Fragment 0:0
> [Error Id: 13e7a786-1135-410f-a4f0-877eab9222d6 on maprdemo:31010]
> ]----
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
