Custom Query (2297 matches)

I have a membase / couchbase cluster running with 3 nodes. I use it to serve images. I moved it to membase for scalability, it easily brings new nodes online, rebalances the data troughout all the nodes.

I noticed 2 things:

1. the nginx config does support multiple servers, but only as fallback.
2. If you configure a fallback when the key doesn't exists, it requests it from both servers even if the key does exist

I would like to submit a feature request that nginx automatically requests the data at the correct membase shard. More information can be found here: ​http://www.couchbase.com/develop

Thanks!

While testing application template for Virtuozzo (nginx-ubuntu-11.10-x86-ez-4.0.0-1.noarch.rpm):

21.03.2012-20.46.25: Selecting previously deselected package nginx. 21.03.2012-20.46.25: (Reading database ... 22168 files and directories currently installed.) 21.03.2012-20.46.25: Unpacking nginx (from .../nginx_1.0.14-1~oneiric_i386.vz.deb) ... 21.03.2012-20.46.25: ---------------------------------------------------------------------- 21.03.2012-20.46.25: 21.03.2012-20.46.25: Thanks for using NGINX! 21.03.2012-20.46.25: 21.03.2012-20.46.25: Check out our community web site: 21.03.2012-20.46.25: * ​http://nginx.org/en/support.html 21.03.2012-20.46.25: 21.03.2012-20.46.25: If you have questions about commercial support for NGINX please visit: 21.03.2012-20.46.25: * ​http://www.nginx.com/support.html 21.03.2012-20.46.25: 21.03.2012-20.46.25: ---------------------------------------------------------------------- 21.03.2012-20.46.26: Setting up nginx (1.0.14-1~oneiric) ... 21.03.2012-20.46.26: nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address already in use) 21.03.2012-20.46.27: nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address already in use) 21.03.2012-20.46.27: nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address already in use) 21.03.2012-20.46.28: nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address already in use) 21.03.2012-20.46.28: nginx: [emerg] bind() to 0.0.0.0:80 failed (98: Address already in use) 21.03.2012-20.46.29: nginx: [emerg] still could not bind() 21.03.2012-20.46.29: invoke-rc.d: initscript nginx, action "start" failed. 21.03.2012-20.46.29: dpkg: error processing nginx (--configure): 21.03.2012-20.46.29: subprocess installed post-installation script returned error exit status 1 21.03.2012-20.46.29: Errors were encountered while processing: 21.03.2012-20.46.29: nginx 21.03.2012-20.46.29: E: Sub-process /usr/bin/dpkg returned an error code (1) 21.03.2012-20.46.29: Error: /usr/bin/apt-get failed, exitcode=100

It looks that the same is now for ubuntu 10.04 (while it had worked before)

Running ssi.t with

    postpone_output 0;

added results in the following alert:

... [alert] ... *18 the http output chain is empty while sending to client ...

Looks like a problem in write filter, it doesn't handle sync flag in buffer correctly. Needs detailed investigation.