I've had Nginx & Thin playing nicely on Ubuntu 10.04 for a couple of years. I have a script in /etc/init.d that starts Thin, using some .yml files in /etc/thin.
All of a sudden restarting Thin for any reason causes a permission denied error in Nginx.
Thin is running on a couple of sockets in /tmp, owned by root. I don't know if that was always the case...but if I change the owner of the socket to www-data (used by Nginx) everything works fine. So I wrote another script to handle it.
Thin is running v1.2.7, and Nginx is 0.7.65. Neither has changed. I did upgrade to Ruby 1.9.3p194 from 1.9.2p290.
Does anybody else have this issue, and have you found a more reasonable fix than "manually" updating permissions?
Upgrading Thin to 1.3.1 solved it. I should've tried that right away, but I had issues with a staging server.
Anyway. Maybe this will help someone else.
Related
I'm new to concourse and really excited to start working with it but I have a problem running the hello world example described here: https://concourse-ci.org/hello-world.html
I'm running this example on a concourse docker setup described here: https://concourse-ci.org/docker-repository.html.
Everything seems to work just fine but when I want to verify the results of both examples it keeps saying loading:
Task result loading (image)
Any idea why this would happen? I'm running docker-compose on Mac OS X (El Capitan) but that shouldn't matter right? Is there some additional configuration that I'm missing?
I also noticed when checking the network trace that the following request doesn't return any value: /api/v1/builds/<buildnumber>/events
It keeps saying 'pending'. Is that normal? I assume it isn't but I don't know the cause of this. Is there any logging I can check?
EDIT:
It seems to have something to do with the fact that it isn't running on localhost. When I use port forwarding and open concourse on localhost:8080 the results are shown just fine. Also mapping another hostname to 127.0.0.1 with port forwarding enabled works. So only when I communicate directly with the opened docker ports it doesn't work. Am I missing something?
After much frustration I found out that to cause of this issue was that Sophos Anti-Virus was blocking Concourse server-side events...
https://community.sophos.com/products/free-antivirus-tools-for-desktops/f/sophos-anti-virus-for-mac-home-edition/5750/sophos-av-blocks-server-sent-events-sse-on-mac-os-x-yosemite
I'm trying to install Grafana on a server, and installation goes through properly. However, when I try to start the service (using sudo service grafana start) it fails with the cryptic message:
2016/02/11 18:45:38 [web.go:93 StartServer()] [E] Fail to start server: open : no such file or directory
I have been unable to find an answer to this.
I assume that I'm simply missing an apt-get package or something really simple, but there's no more information than this.
Anyone have an idea?
Thanks for your time.
EDIT:
While unable to solve the actual problem, I realized that though I configured the server to run over HTTPS, the actual SSL is handled through the proxy by my host, and the server should run internally on HTTP. When changing this, the server started properly. It's not a solution to this specific problem, but as it may point others with this problem in the right direction;
the problem had to do with running over HTTPS.
Good luck!
when configuring Grafana to use HTTPs you need to specify cert & key paths, looks likely that Grafana could not find one of them.
Installed this (https://github.com/BanzaiMan/openshift-origin-cartridge-phppgadmin) to my Tiny Tiny RSS application on OpenShift to manage my database.* However, after the installation and a few restarts, .../phppgadmin/ URL gives me 404 error. Any ideas? Could it be the github cartridge is using old environmental variables? Thanks!
*The reason I want to install phppgadmin in the first place is to vacuum my ever-expanding database on Tiny Tiny RSS application. vacuumdb and vacuumdb -f -a only claim ~50mb and app-tidy does ~100mb, as opposed to ~600mb previously. So, I need to find another solution, like phppgadmin, to address my quota limitations.
Instead of using phppgadmin, it would be a much better idea to download pgadmin onto your computer and then use the rhc port-forward command to connect to your database and do your maintenance. That cartridge is pretty old and likely will not get updated anytime soon (or ever).
I've followed instructions online to set up a Telescope instance on my DigitalOcean droplet, but it won't start with Upstart.
I'm able to run the server successfully manually, but the Upstart task doesn't fire when the server boots. I'm sure I should be looking at a log file somewhere to discover the problem, but I'm not sure where.
I've looked for the location of upstart logs, but I'm not having any luck. Either you have to add something to your script to make it log, or it just does it according to accounts online, but neither of those seem to be the case for me.
When I try to search for help on Upstart, I'm also seeing people saying I should be using systemd instead, but I can't figure out how to install it on CentOS 6.5.
Can anyone help me figure a way out of this labyrinth?
I use Ubuntu server 14.04, and my upstart logs are located in /var/log/upstart
The log usually contains stdout from the job, and it should help you understand what's wrong.
My guess is that when the server boots and tries to run your job, MongoDB is not yet ready so it fails silently.
Try installing the specific MongoDB version that Meteor is using at the moment (2.4.9) using these docs :
http://docs.mongodb.org/v2.4/tutorial/install-mongodb-on-ubuntu/
The most important thing is to get upstart support for MongoDB, this will allow us to catch mongod launch as an event.
You can then use this syntax in your upstart script :
start on started mongodb
This will make your node app start when mongo is ready.
I've created a gist with the scripts I wrote to setup a server ready for Meteor app deployment, it's a bit messy and probably specific to Ubuntu but it might help you.
https://gist.github.com/saimeunt/4ace7975b12df06ee0b7
I'm also using demeteorizer and forever which are two great tools you should probably check.
CentOS 5.5
I have a web application running on a server and it needs access to another CentOS server's file system running in the same network (via private IP). After doing a bunch of googling it looks like mounting the drive via NFS is a good way to go, but I'm not finding any good step by step instructions on how to go about it. I've read the man docs on the mount command and read some docs on the CentOS wiki as well but I feel like I'm missing something. Here is what I'm trying
mount -t nfs my.ip.address:/somePath /somePath/mount
I keep getting a 'no route to host' error but I can ping the server just fine. I'm guessing that I am possibly missing a port I need to open or something, but again, can't find information that makes sense to a non-sysadmin like myself.
Thanks for any help.
I ran across this, followed it step by step, and now I'm up and running!
http://www.cyberciti.biz/faq/centos-fedora-rhel-nfs-v4-configuration/