Rsync to copy Files Between Servers and Computers

Posted on June 28, 2014 by curiouscat

I found rsync when I wanted to use scp to copy files to a server but not overwrite files already there. Rsync is actually more efficient no matter what (it seems) but it is really great if there are a bunch of duplicate files (Rsync will just skip them).

To copy files from your current computer to a server:

rsync -azv -r ssh [directory] [username]@remotehost:[location]

[bash]$ sudo rsync -azv -e ssh directory-on-my-computer/ [email protected]:~/files/go/here[/bash]

a = archive mode
z = compress file data
v = verbose
r = recurse through subdirectories and copy all of them

Compressing file data saves bandwidth so if that is an issue it is another big win over scp. And in my reading it seems rsync can restart a broken file transfer in process (while scp you have to redo the whole file transfer).

To copy from the server to your computer just reverse the order of the locations. And you can even just put in two addresses not on your current computer and copy between then.

As a reminder, I realize this blog is made up of stuff that is obvious to a large number of people. It is really aimed at me (so I can quickly find what I found before), and to a lessor extent others like me (who use cli some but are not system administrators or programmers to any significant extent).

Bad iTunes Ux and How to Submit a Podcast to iTunes

Posted on May 21, 2014 by curiouscat

I find iTunes Ux to be horrible. My guess it is largely driven by people that think the useful web isn’t cool enough so lets break all the things that work in order to be different. Or maybe there is some other reason for the horrible Ux that I don’t understand.

iTunes doesn’t use urls which are super user friendly. And even when urls are lamely named they can be useful.

Trying to figure our using the UI of iTunes how to submit a podcast was impossible. using the iTunes help found nothing for submit podcast? Using Google to overcome Apple’s lame help found web pages on Apple telling you what to do. It read like 1990 UI navigation by text. Go here click here, then click, here then click here. How idiotic. On top of being idiotic and outdated it didn’t work the places to click didn’t exist.

With using 1997 technology (way before really but wasn’t so common long before 1997) just have

itunes.com/podcasts/submit

No help needed. Even if someone goes to help just give that clickable url and done. But no, in 2014 Apple still can’t do that.

Finally I found another url

https://buy.itunes.apple.com/WebObjects/MZFinance.woa/wa/publishPodcast

Super lame and unreadable but guess what clicking it works, unlike the various useless Apple webpages on clicking here and then clicking there and then clicking the next place which doesn’t exist.

Turn It Off and On Again

Posted on May 8, 2014 by curiouscat

I had a WordPress blog lose the database connection. As I tried to troubleshoot it I rapidly got to the point of thinking that maybe just rebooting the server would fix things – since no changes had been made that should cause the database connection to be lost. But also I figured I couldn’t be so lucky that such a simple thing would work.

I tried to log into the database using the cli and it was failing. MySQL was still running – according to top. I have automated security updates setup for the VPS server running Ubuntu. My thought now is those updates somehow messed things up to the extent the server database connections somehow wasn’t working. I sure hoped that was the case.

I rebooted and hoped.

After a bit I was able to see that it was working.

When I went to reboot I noticed I hadn’t rebooted in a long time. My guess is I might have the automated security updates setup wrong on this server (maybe it is installing more than just the security updates), I’ll check. I have not had this problem before with other servers, obviously breaking a database in this way would cause lots of problems on production machines so I tend to think it is more likely I messed something up than this is a likely outcome when using automated security updates (but such things are possible which is why I think places with full time system admins and important servers rely on manual updates with professionals watching everything to be sure nothing obvious breaks).

My feeling is the longer you go with not rebooting the server the more likely some issues are to crop up (but also every reboot is more like to result in some broken thing, right then, than doing nothing so it is a tradeoff). And I could be wrong in that feeling, it is just what I guess without much evidence to support my guess.

Anyway I was very happy turning it off and on again worked. Honestly I went ahead and updated the server before I rebooted, but my guess is just turning it off and on again would have worked.

Top with Better Display Options

Posted on March 10, 2014 by curiouscat

Scout Realtime is a Ruby Gem that allows you to view top in the browser. One huge advantage is to view charts of activity over time.

Scout Realtime is open source and free.

Make Your Bounce Rate More Meaningful

Posted on May 26, 2013 by curiouscat

Nice post on what the operational definition (though it doesn’t use that term) of bounce rate is in Google Analytics and some ideas on how to fix bounce rate to be more meaningful (better fit your desired operational definition).

We often make mistakes because we use the data we have assuming it means something other than what it actually means. There suggestions let you make Google Analytics collect the data so it is more aligned to what you want to know.

Essentially bounce rate without any adjustment is operationally defined by Google to mean the visitor doesn’t click a link to another page on your site that has Google Analytic code (so it can register the click).

They suggested using

[javascript]setTimeout("_gaq.push([‘_trackEvent’, ’15_seconds’, ‘read’])", 15000);[/javascript]

inside your Google analytic tracking code. They also updated the post with an idea from a Hacker News comment to use scroll tracking as a better option than just timing if they stay on the page at least 15 seconds. The scroll tracking lets you count it as a bounce only if they don’t scroll down and they don’t click to a new page on your site.

Which data is more useful depends on your situation, what you want to know and what you want to do with the data you collect.

Don’t Copy-Paste Directly from Website to Terminal

Posted on April 8, 2013 by curiouscat

Good reminder and discussion on Hacker News about the danger of copy-paste from a website into terminal.

When copying from a website it is easy for hidden text to be included in the clipboard. If you paste that into terminal it can be executed before you see the code. Including new line commands in the copied text will automatically execute the commands (in most terminal applications).

The best solution is to paste the clipboard content into something to view the text before pasting into clipboard. The best way is probably to recopy it. I use this double copy method to get rid of formatting I don’t want (when the clipboard includes things like font styling info you don’t want – not relavent when pasting into terminal but the pasting [into say vim, which won’t include the extra formatting details] and recopying part is similar). It may be possible to have text hidden (write it in a way where it won’t show up on the screen but is in the clipboard – using some tricky unicode characters or something).

Some people put # before pasting into clipboard but that only protects the first “line.” Any new lines could still be run without you seeing them.

The discussion warns against a malicious website intentionally creating a problem but if you paste in multiple lines there is also just the danger from the lines executing immediately when you wanted to edit the line before it was executed.

Making Sure You Don’t Run Out of Space on Your VPS

Posted on February 6, 2013 by curiouscat

Over the years I have had web sites I work on go down because the server ran out of space (when I wasn’t responsible for the system administration). Well today I had it happen for some sites on one of my servers.

I have a remote monitoring service so I was notified when the sites were unavailable. Looking in the error logs I could see I ran out of space, which is a fairly lame error to make, I think. I cleared up some space quickly and things were in working order again.

A few tips from this experience. To see data on space used and available space from the command line (linux/ubuntu) use the df command.

[bash]df[/bash]

You could check that occasionally to see if anything strange is going on. But it is better to use something like Nagios to monitor the server and provide notice if usage goes past certain points is wise (75%, 80%…).

I found a nice way to find large files (that may be possible to delete to free up space)

[bash]sudo find . -mount -ls | awk ‘{print $7, $11}’ | sort -rn > large_file_list[/bash]

If you get an error with the file creation, you can log in as root (su) or create the file first and then it can overwrite the existing file.

It is smart to have log rotate setup for logs. I didn’t have it setup for every log. To create a file to keep track of all your logs for domains you can do the following. Create a file in the logrotate directory so it is run per your settings.

[bash]sudo vim /etc/logrotate.d/virtualhosts[/bash]

This article from Slicehost provide details on options, a simple setup is to include text such as this in the file created in the line above.

[bash]/var/www/domain1.com/logs/*log /var/www/domain2.com/logs/*log {
rotate 14
daily
compress
delaycompress
sharedscripts
postrotate
/usr/sbin/apache2ctl graceful > /dev/null
endscript
}[/bash]

Continue reading →

Bash Profile Adjustments – Scrolling History

Posted on January 28, 2013 by curiouscat

Your bash profile lets you make edits that let you design how the cli works for you. One of my favorite edits is to let me set the scrolling back through my history be based on what I start typing. So, for example, I can type

[bash]cd[/bash]

and then use the up-arrow key and it scrolls through my cli history based only on those that start with cd

To add this to your .profile file (or put it in your .bash_rc file):

[bash]bind ‘"\e[A"’:history-search-backward
bind ‘"\e[B"’:history-search-forward[/bash]

ModSecurity – Adjustments for WordPress

Posted on January 6, 2013 by curiouscat

How to secure an Ubuntu 12.04 LTS server – Part 1 The Basics is a great resource to secure a VPS.

Mod_Security is a web application firewall. I found a couple things to add to my servers.

The same site includes a very good guide to installing it (you also may well want to whitelist Googlebot, instructions in the link). However it blocked my access to one of my sites. You end up just getting the message:

“Forbidden

You don’t have permission to access / on this server.”

There are tips on some edits to include when using WordPress on an Apache server with modsecurity.

One simple action is to include

SecRule SERVER_NAME “[your-domain.com]” phase:1,nolog,allow,ctl:ruleEngine=off

[bash]SecRule SERVER_NAME "code.curiouscatnetwork.com" phase:1,nolog,allow,ctl:ruleEngine=off[/bash]

on the bottom of modsecurity.conf which is found /etc/modsecurity (for me, on Ubunutu 12.04). Then restart Apache

[bash]sudo service apache2 restart[/bash]

and see if the problem goes away. If it does then you have a very good indication modsecurity was blocking access and can continue to narrow the scope of the problem by adding the WordPress whitelist rules in the link above.

Another note, service apache2 start, failed in a non-obvious way to me anyway. For me if I use sudo it works fine. If I don’t it gives odd errors which lead me on a 10 minute wild goose chase before remembering to try sudo.

How To Transfer Domain Registrar From Godaddy To Gandi.net

Posted on November 8, 2012 by curiouscat

I was asked to take over managing a few domains for someone, part of this includes renewing the registrations. As they currently are on Godaddy and having to deal with that system would annoy me and waste my time I am transferring the registrar to Gandi.net.

As part of the transfer process you must make sure the domain is unlocked, which I did. Then you must send yourself the authorization code (for making a transfer). You need to provide this code to the new registrar. Not surprisingly Godaddy attempts to make this difficult. Then the email just says: prog_id=GoDaddy&isc=aaaa111 That will not work. I was able to find a way to download the code for all the domains (you have to create an export list, instead of just viewing it from the domain page). That method provided a 16 digit (alpha and numeric and very occasional special character) codes for all the domains.

Gandi.net makes the transfer process very nice and smooth with a easy to follow user interface. They do cost a bit more than others but they are well worth it in the time you save for most people (if you really want to save every last dollar you can find cheap registrars that are less annoying than Go Daddy).