Home / server / Questions / 370531
Accepted
Michael
Asked: 2012-03-17 11:09:52 +0800 CST

Mongo Client RedHat EL5 UT8 Support

  • 772
# mongo
MongoDB shell version: 1.6.4
Fri Mar 16 11:55:46 *** warning: spider monkey build without utf8 support. consider rebuilding with utf8 support
connecting to: test

Mongo Server seems to handle the utf8 characters fine, as well as my php-mongo-client driver.

But when I try to query a record that has a utf8 character from the mongo command line client I get:

> db.Users.find({age:33});
error:non ascii character detected
Fri Mar 16 11:55:43 mongo got signal 11 (Segmentation fault), stack trace: 

Fri Mar 16 11:55:43 0x440b50 0x3664c302d0 0x3f47e7b6e0 0x3f47e83bbd 0x3f47e254f3 0x3f47e25660 0x3f47e256ee 0x3f47e25792 0x3f47e2876e 0x4b031d 0x443b72 0x445476 0x3664c1d994 0x43fd39
 mongo(_Z12quitAbruptlyi+0x3b0) [0x440b50]
 /lib64/libc.so.6 [0x3664c302d0]
 /usr/lib64/libjs.so.1 [0x3f47e7b6e0]
 /usr/lib64/libjs.so.1(js_CompileTokenStream+0x3d) [0x3f47e83bbd]
 /usr/lib64/libjs.so.1 [0x3f47e254f3]
 /usr/lib64/libjs.so.1(JS_CompileUCScriptForPrincipals+0x60) [0x3f47e25660]
 /usr/lib64/libjs.so.1(JS_EvaluateUCScriptForPrincipals+0x3e) [0x3f47e256ee]
 /usr/lib64/libjs.so.1(JS_EvaluateUCScript+0x22) [0x3f47e25792]
 /usr/lib64/libjs.so.1(JS_EvaluateScript+0x6e) [0x3f47e2876e]
 mongo(_ZN5mongo7SMScope4execERKSsS2_bbbi+0xed) [0x4b031d]
 mongo(_Z5_mainiPPc+0x14a2) [0x443b72]
 mongo(main+0x26) [0x445476]
 /lib64/libc.so.6(__libc_start_main+0xf4) [0x3664c1d994]
 mongo(__gxx_personality_v0+0x269) [0x43fd39]

Any ideas or suggestions would be welcome

redhat utf-8 mongodb

3 Answers

  1. Best Answer

    The mongo version 1.6.4 was installed by the package management system, which includes the mongo init scripts.

    The solution I came up with was to download the newest Mongo(2.0.3) binaries from http://www.mongodb.org/downloads. Then I replaced all the binaries from the rpm /usr/bin with the ones I downloaded.

    Everything now works perfectly!

    • 3

  2. Just had the same issue. MongoDB shell version: 1.8.2 on Ubuntu (EC2)

    > db.tweets.find();

    error:non ascii character detected

    Checked which version of MongoDB I was running by issuing the command:

    $ mongod --version

    db version v1.8.2, pdfile version 4.5 - Wed Oct 24 15:43:13 git version: nogitversion

    Confirm if mongo is currently running:

    $ ps -deaf | grep mongod

    mongodb 15408 1 0 Jun06 ? 13:50:00 /usr/bin/mongod --config /etc/mongodb.conf

    To Shutdown MongoDB

    $ ./mongo

    > use admin

    > db.shutdownServer()

    server should be down...

    Then upgraded following these instructions: How to Install MongoDB on Ubuntu

    Add the 10 Gen (creators of MongoDB) public key to apt-get so you trust their package:

    $ sudo apt-key adv --keyserver keyserver.ubuntu.com --recv 7F0CEB10

    $ echo "deb http://downloads-distro.mongodb.org/repo/ubuntu-upstart dist 10gen" >> /etc/apt/sources.list.d/10gen.list

    Update your packages:

    $ sudo apt-get update

    install 10gen's MongoDB Debian/Ubuntu package:

    $ sudo apt-get install mongodb-10gen

    The following packages will be REMOVED: mongodb

    The following NEW packages will be installed: mongodb-10gen

    If you get an error, try:

    $ apt-get autoremove

    once you have the new version of MongoDB the "non ascii" character issue should be gone.

    (p.s. I know this question relates to RHEL but people running other versions of Linux will come up against this issue so I thought a Ubuntu/Debian answer was relevant. Feel free to remove

    • 1

  3. Take a look at:

    http://www.mongodb.org/display/DOCS/Building+Spider+Monkey#BuildingSpiderMonkey-Build

    "SpiderMonkey does not use UTF-8 by default, so we enable before building."

    So, you'll need to rebuild SpiderMonkey.

    • 0