Book Excerpt: DevOps Troubleshooting: Linux Server Best Practices
What you will notice as you refresh this page is that the objects in the scoreboard should change during each request. This scoreboard is handy when you want to keep an eye on your web server; just continually refresh the page. During a spike, you are able to see new processes get spawned, switch to W to serve the request, and then, if the spike in traffic subsides, those processes slowly change to _ and ultimately . as they are no longer needed.
Generally speaking, when you access the server status page, you do so from the command line while logged into the web server. This lets you restrict what hosts can view the page while still providing all the information you need. Now, by default, if you were to run curl against the regular server status page, you would get HTML output. However, if you pass the auto option to the server-status page, you will get text output that’s more useful for both command-line viewing and parsing by scripts:
$ curl http://localhost/server-status?auto % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 117 586 117 586 0 0 2579 0 --:--:-- --:--:-- --:--:-- 2579117 586 11 7 586 0 0 1905 0 --:--:-- --:--:-- --:--:-- 0 Total Accesses: 2343235 Total kBytes: 265925501 CPULoad: .0773742 Uptime: 6801454 ReqPerSec: .34452 BytesPerSec: 40036.7 BytesPerReq: 116210 BusyWorkers: 53 IdleWorkers: 28 Scoreboard: WW_W__W_W__W_K_W_W_K___WWWWW_WWKWWWW_WWWWWWWWWWWWWWWWWWKKKKKK_KW_.WC.CW_____K__ ____....................................................................................... ........................................................................................... ........................................................................................... ................................................
When you want to monitor the status page of a server in the command line, although you could just run the curl command over and over, you could use a handy command called watch, which will run whatever command you specify every x number of seconds (by default 2). So if you wanted to keep an eye on the status page and have it refresh every 5 seconds on the command line, you would type
$ watch -n 5 'curl http://localhost/server-status?auto'
To exit out of watch, just hit Ctrl-C.
Solve Common Web Server Problems
Although it’s difficult to document how to solve any and all web server problems, you are likely to run into a few general classes of problems that have identifiable symptoms. This section will highlight some of the common types of problems you may find, their symptoms, and how to remedy them.
One common and relatively simple problem to identify in a web server is a configuration problem. Since web servers need to be reloaded to take on changes in their configuration, it can be tempting to make many changes without reloading the web server; however, if you do so, you are likely to find out during a server maintenance (or when you need to restart the server to load new SSL certificates) that there’s some syntax error in your configuration files and your server will refuse to start.
Both Apache and Nginx validate their configuration files when you start, restart, or reload the service, so that’s one way to find configuration errors—unfortunately, it also means that in the case of a problem, the server is down while you fix the errors. Fortunately, both web servers provide means to test configuration syntax and highlight any syntax errors while the server is still running.
In the case of Apache, the command is apache2ctl configtest. Be sure to run this command as a user who can read all of the configuration files (probably the root user). A successful run looks like this:
$ sudo apache2ctl configtest Syntax OK
When there is a syntax error, this command will identify the file and line number of the error so it’s easy to find:
$ sudo apache2ctl configtest apache2: Syntax error on line 233 of /etc/apache2/apache2.conf: Could not open configuration ↪file /etc/apache2/conf/: No such file or directory
In this case, the configuration file had a typo—the directory you wanted to include was /etc/apache2/conf.d.
Nginx also provides a syntax check by running nginx -t:
$ sudo nginx -t the configuration file /etc/nginx/nginx.conf syntax is ok configuration file /etc/nginx/nginx.conf test is successful
As with Apache, when Nginx detects an error, it tells you the file and line number:
$ sudo nginx -t [emerg]: unknown directive "included" in /etc/nginx/nginx.conf:13 configuration file /etc/nginx/nginx.conf test failed
A common headache, especially for new web server administrators, is permission problems on the web server. Although both Apache and Nginx’s initial processes run as root, all subprocesses that actually do the work of serving content run as a user with more restricted permissions—usually a user like www-data or apache. If you are, for instance, uploading web pages as a different user, you may initially run into permissions problems until you make sure that each file you want to serve is readable by the www-data or apache user.
So what does a permissions problem look like from the outside? This example takes a basic Nginx setup and changes the permissions on the main index.html file so that it is no longer readable by the world. Then it uses curl to attempt to load the page:
$ curl http://localhost <html> <head><title>403 Forbidden</title></head> <body bgcolor="white"> <center><h1>403 Forbidden</h1></center> <hr><center>nginx/0.7.65</center> </body> </html>
The output from the web page tells us the HTTP error even without having to tell curl to display it: a 403 Forbidden error. Unfortunately, although we can see the page is forbidden, from this output, we’re not yet sure why. At this point, though, we would turn to the Nginx error logs and see
2012/07/07 16:13:37 [error] 547#0: *2 open() "/var/www/nginx-default/index.html" failed ↪(13: Permission denied), client: 127.0.0.1, server: localhost, request: "GET / ↪HTTP/1.1", host: "localhost"
This error log lets us know that Nginx attempted to open /var/www/nginx-default/index.html, but permission was denied. At this point, we could check out the permissions of that file and confirm that it wasn’t readable by the www-data user Nginx runs as:
$ ps -ef | grep nginx root 545 1 0 15:19 ? 00:00:00 nginx: master process /usr/sbin/nginx www-data 547 545 0 15:19 ? 00:00:00 nginx: worker process $ ls -l /var/www/nginx-default/index.html -rw-r----- 1 root root 151 2006-08-30 03:39 /var/www/nginx-default/index.html
In this case, you could fix this permission problem with the chmod o+r command, which would add world read permissions to the file. Alternatively, you could also change the group ownership of the file so it was owned by the www-data group (or by a group that www-data was a member of).
Although some administrators may sidestep permissions problems by basically making all files readable and writeable by everyone, the security risks of doing so aren’t worth the easy fix. Instead, consider creating a group on the system whose membership includes both www-data or apache users (depending on what user your web server runs as) and the users you upload files as. If you do try the “chmod 777” method of making the file readable by everyone, use it only as a temporary sanity check to confirm that the problem truly is a permissions issue. Be sure after you have solved the problem to change permissions back to something more secure.
Kyle Rankin is a VP of engineering operations at Final, Inc., the author of a number of books including DevOps Troubleshooting and The Official Ubuntu Server Book, and is a columnist for Linux Journal. Follow him @kylerankin.
Fast/Flexible Linux OS Recovery
On Demand Now
In this live one-hour webinar, learn how to enhance your existing backup strategies for complete disaster recovery preparedness using Storix System Backup Administrator (SBAdmin), a highly flexible full-system recovery solution for UNIX and Linux systems.
Join Linux Journal's Shawn Powers and David Huffman, President/CEO, Storix, Inc.
Free to Linux Journal readers.Register Now!
|CentOS 6.8 Released||May 27, 2016|
|Secure Desktops with Qubes: Introduction||May 27, 2016|
|Chris Birchall's Re-Engineering Legacy Software (Manning Publications)||May 26, 2016|
|ServersCheck's Thermal Imaging Camera Sensor||May 25, 2016|
|Petros Koutoupis' RapidDisk||May 24, 2016|
|The Italian Army Switches to LibreOffice||May 23, 2016|
- Secure Desktops with Qubes: Introduction
- Download "Linux Management with Red Hat Satellite: Measuring Business Impact and ROI"
- CentOS 6.8 Released
- The Italian Army Switches to LibreOffice
- Linux Mint 18
- Chris Birchall's Re-Engineering Legacy Software (Manning Publications)
- ServersCheck's Thermal Imaging Camera Sensor
- Petros Koutoupis' RapidDisk
- Oracle vs. Google: Round 2
- The FBI and the Mozilla Foundation Lock Horns over Known Security Hole
Until recently, IBM’s Power Platform was looked upon as being the system that hosted IBM’s flavor of UNIX and proprietary operating system called IBM i. These servers often are found in medium-size businesses running ERP, CRM and financials for on-premise customers. By enabling the Power platform to run the Linux OS, IBM now has positioned Power to be the platform of choice for those already running Linux that are facing scalability issues, especially customers looking at analytics, big data or cloud computing.
￼Running Linux on IBM’s Power hardware offers some obvious benefits, including improved processing speed and memory bandwidth, inherent security, and simpler deployment and management. But if you look beyond the impressive architecture, you’ll also find an open ecosystem that has given rise to a strong, innovative community, as well as an inventory of system and network management applications that really help leverage the benefits offered by running Linux on Power.Get the Guide