通过TCP在单个数据包中传输数据

Question

如果CSV文件中的行大约为10或2万（1或2 MB），则此脚本按预期工作。如果我尝试推送10万行（10 MB），我会收到错误。

import elasticsearch

es = elasticsearch.Elasticsearch('https://some_site.com/')  

import csv
import urllib2
response=open('/var/lib/mysql_orig/test/head.csv', 'r')
csv_file_object = csv.reader(response)

header=['s_id', 'col1', 'col2', 'col3', 'col4', 'col5', 'col6', 'col7', 'col8', 'col9', 'col10', 'col11']
bulk_data = [] 
op_dict=dict()
for row in csv_file_object:
    #row = unicode(row, errors='ignore')
    data_dict = {}
    op_dict={}
    for i in range(len(row)):
        data_dict[header[i]] = unicode(row[i], errors='ignore')
        op_dict = {
            "index": {
                 "_index": "student", 
                "_type": "marks", 
                "_id": data_dict['s_id']
            }
        }
    bulk_data.append(op_dict)
    bulk_data.append(data_dict)

es.indices.delete(index =  'student')
es.indices.create(index =  'student', body = request_body)
# bulk index the data
res = es.bulk(index =  'student', body = bulk_data, refresh = True)

ConnectionError: ConnectionError(('Connection aborted.', error(32, 'Broken pipe'))) caused by: ProtocolError(('Connection aborted.', error(32, 'Broken pipe')))

是否有任何设置允许我通过TCP发送大数据包（默认弹性搜索端口9200）？