You are viewing a plain text version of this content. The canonical link for it is here.
Posted to oak-issues@jackrabbit.apache.org by "Marcel Reutegger (JIRA)" <ji...@apache.org> on 2018/12/13 09:14:00 UTC

[jira] [Created] (OAK-7959) MongoDocumentStore causes scan of entire nodes collection on startup

Marcel Reutegger created OAK-7959:
-------------------------------------

             Summary: MongoDocumentStore causes scan of entire nodes collection on startup
                 Key: OAK-7959
                 URL: https://issues.apache.org/jira/browse/OAK-7959
             Project: Jackrabbit Oak
          Issue Type: Bug
          Components: mongomk
    Affects Versions: 1.9.13, 1.9.12, 1.9.11, 1.9.10
            Reporter: Marcel Reutegger
            Assignee: Marcel Reutegger


This is a regression introduced with OAK-7645 when the MongoDB Java driver was updated to 3.8.

With the switch to the new driver, the use of a deprecated method {{count()}} was replaced with {{countDocuments()}}. The method {{countDocument()}} behaves differently and performs an aggregation command over all documents in the collection. The MongoDB log would then show something like:
{noformat}
2018-12-12T18:40:53.672+0100 I COMMAND  [conn6063] command oak.nodes command: aggregate { aggregate: "nodes", readConcern: { level: "local" }, pipeline: [ { $match: {} }, { $group: { _id: null, n: { $sum: 1 } } } ], cursor: {}, $db: "oak", $readPreference: { mode: "primaryPreferred" } } planSummary: COLLSCAN keysExamined:0 docsExamined:4038809 cursorExhausted:1 numYields:31584 nreturned:1 reslen:127 locks:{ Global: { acquireCount: { r: 31586 } }, Database: { acquireCount: { r: 31586 } }, Collection: { acquireCount: { r: 31586 } } } protocol:op_msg 1642ms
{noformat}




--
This message was sent by Atlassian JIRA
(v7.6.3#76005)