partial update 原理 和 优点
es,其实是有个内置的脚本支持的,可以基于groovy脚本实现各种各样的复杂操作
基于groovy脚本,如何执行partial update
es scripting module,后面会详细讲解,这里就只是初步讲解一下
PUT /test_index/test_type/11{ "num": 0, "tags": []}
(1)内置脚本
POST /test_index/test_type/11/_update{ "script" : "ctx._source.num+=1"}{ "_index": "test_index", "_type": "test_type", "_id": "11", "_version": 2, "found": true, "_source": { "num": 1, "tags": [] }}
(2)外部脚本
ctx._source.tags+=new_tag
POST /test_index/test_type/11/_update{ "script": { "lang": "groovy", "file": "test-add-tags", "params": { "new_tag": "tag1" } }}
(3)用脚本删除文档
ctx.op = ctx._source.num == count ? 'delete' : 'none'
POST /test_index/test_type/11/_update{ "script": { "lang": "groovy", "file": "test-delete-document", "params": { "count": 1 } }}
(4)upsert操作
POST /test_index/test_type/11/_update{ "doc": { "num": 1 }}{ "error": { "root_cause": [ { "type": "document_missing_exception", "reason": "[test_type][11]: document missing", "index_uuid": "6m0G7yx7R1KECWWGnfH1sw", "shard": "4", "index": "test_index" } ], "type": "document_missing_exception", "reason": "[test_type][11]: document missing", "index_uuid": "6m0G7yx7R1KECWWGnfH1sw", "shard": "4", "index": "test_index" }, "status": 404}如果指定的document不存在,就执行upsert中的初始化操作;如果指定的document存在,就执行doc或者script指定的partial update操作
POST /test_index/test_type/11/_update{ "script" : "ctx._source.num+=1", "upsert": { "num": 0, "tags": [] }}
partial update 原理
retry策略
再次获取document数据和最新版本,再次更新
post /index/type/id/_update?retry_on_conflict=5&version=6