@anarasimham Thanks for quick reply The reason I was using mergeRecord with defragment was that the requirement is to keep all the nested json object extracted from a file to be kept in one output file. Using filename as correlation attribute, the problem is: 1. using defragment algo: I don't know upfront the size of json object extracted from nested arrays, which I could then use as fragment.count 2. using bin-packing algo: Again I don't know the predictable minimum time it will take to complete processing an input file, which I could use as bin-age I have multiple json each with below schema: [
{
"field1":"value1",
"nested-array":[
{
"inner-field1":"inner-value1",
"inner-field2":"inner-value2",
...
},
{
"inner-field1":"inner-value3",
"inner-field2":"inner-value4",
...
},
{
"inner-field1":"inner-value5",
"inner-field2":"inner-value6",
...
}
]
},
{
"field1":"value1",
"nested-array":[
{
"inner-field1":"inner-value7",
"inner-field2":"inner-value8",
...
},
{
"inner-field1":"inner-value9",
"inner-field2":"inner-value10",
...
}
]
}
] I need to write nested-array object from ALL json files to a single csv file, like below "inner-field1","inner-field2",..
"inner-value1","inner-value2",..
"inner-value3","inner-value4",..
"inner-value5","inner-value6",..
"inner-value7","inner-value8",..
"inner-value9","inner-value10",..
...
regards, Hemal
... View more
