[SERVER-1504] Possibly deadlock reached while performing simultaneous queries and mapreduces Created: 28/Jul/10  Updated: 12/Jul/16  Resolved: 28/Jul/10

Status: Closed
Project: Core Server
Component/s: Querying
Affects Version/s: 1.5.6
Fix Version/s: 1.5.7

Type: Bug Priority: Major - P3
Reporter: Ken Kurzweil Assignee: Mathias Stearn
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

CentOS Linux 2.6.18-194.8.1.el5 SMP x86_64


Attachments: Text File currentOp_output.txt     Text File mongdb_stdout_stderr.txt    
Issue Links:
Depends
Operating System: Linux
Participants:

 Description   

While performing simultaneous activities I was able to put a mongod 1.5.6 instance into a supposed deadlock. The server still accepts connections but will not perform further queries or mapreduces. I've included a currentOp output and a portion of stdout and stderr from mongod. Samples of the types of mapreduces and queries used can be gleaned from this output. The server was under heavy use with many processes/hosts concurrently querying and inserting and performing mapreduces.

The mongod and the system is idle. I've tried interrupting the server with a response of "got kill or ctrl c or hup signal 2 (Interrupt), will terminate after current cmd ends". I've issued db.killOp requests which do not take effect.



 Comments   
Comment by Mathias Stearn [ 28/Jul/10 ]

fixed in git master

Comment by Ken Kurzweil [ 28/Jul/10 ]

Not using db.eval() and if it helps all interaction is performed by processes using pymongo 1.7.

Comment by Mathias Stearn [ 28/Jul/10 ]

Are you calling map reduce from within a db.eval()?

Generated at Thu Feb 08 02:57:13 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.