Backup sites and databases

[Peter Knight]

Had a discussion with a developer friend this morning about hacking (preventing) and backing up sites and databases.

Here's some of the methods disussed:

• PHP scripts hooking into DropBox/Amazon/Google Drive
• Manually backing up and saving to NAS. 
• Manually backing up, syncing to NAS and then syncing to online service
• Writing CRON jobs which scan the root for file and folder changes, email you alerts and backup at regular intervals.
• Online services such as CodeGuard

What method(s) do you use to backup your sites. I'm talking about any type of site on any CMS.

[adrian]

I have been using these two for a long time. Might be better options out there, as both of these are quite old and haven't been updated in a long time, but they also work great and I am not sure they need updating

http://sourceforge.net/projects/automysqlbackup/ 

http://www.rsnapshot.org/ (incremental backups of the sql backup files and all other site files to a remote server)

[Peter Knight]

Interesting links.

[horst]

this one is rock solid, it is developed since 2003, http://www.mysqldumper.net/features/

[renobird]

I've been using this method Ryan posted. As Ryan mentioned at the end of the post, rsync copies them to my local HD, which is then again backed up via Time Machine. The shell script for backing up the file system is a little more complicated. My version is essentially a tweak of this script on github.

The major drawback is that the backups eat into your server space. I've been meaning to explore the option of moving them Dropbox/Amazon.

So far I have plenty of space, so I'm just letting it ride.

[cstevensjr]

I'm a former Joomla adherent.  I have used paid Akeeba products for years and am now trying their stand-alone (Akeeba Solo) PHP backup solution.  

I've been on vacation for over a month and therefore have not invested too much time into taking this product through it's paces. 

[LostKobrakai]

I'm on a german shared hosting, which is provided by a some linux devs (no marketing/business people). They provided their backup strategy here, but it's written in german. As User I found it super easy to go through backups and restore stuff, that's why I like it and as hostingprovider this has to be bullet prove.

https://wiki.uberspace.de/system:backup

If theres some interest I could maybe translate it. 

[Marty Walker]

All of my sites are on a cPanel VPS so I use the backup to Amazon S3 feature. From there they get shuffled off to Glacier. Previously I used http://www.cpanelbackupscript.com/

[Peter Knight]

Quick followup in case anyone else comes to this thread.

I had a look at all the links and eventually settled on Akeeba Solos (thanks @cstevensjr) as it was the most compatible with my server.

The website is a little unclear but you install it on your server, enter your database credentials and you have quite an array of backup and config options from that point onwards.

Setting up my backups (databases and files) with Dropbox was really simple. There's also 3 different options for scheduled backups via Cron.

[pwired]

My search for this ended with xcloner: http://www.xcloner.com/    standalone version !

This stuff is for real.

[gebeer]

I'm on a german shared hosting, which is provided by a some linux devs (no marketing/business people). They provided their backup strategy here, but it's written in german. As User I found it super easy to go through backups and restore stuff, that's why I like it and as hostingprovider this has to be bullet prove.

https://wiki.uberspace.de/system:backup

If theres some interest I could maybe translate it. 

I'm also using this German hoster for some sites. In a nutshell, they are using rsnapshot for their backups. Their backup server is pulling data from the live server. The backups are mounted via NFS to the live server.

I came from the Joomla universe to PW. Thus I'm familiar with Akeeba backup which is a great product. Using their free and well maintained standalone version makes backing up PW and other PHP applications a breeze.

[mr-fan]

and that this one is not going to hide somewhere....

https://github.com/ryancramerdesign/ProcessDatabaseBackups

(should work great for normal sites)

since it works from the API you could setup a cron or lazycron to get it automatic (maybe send the backupfile via mail to get the db to a second place away from the webspace?)

regards mr-fan

[Christophe]

I saw this topic yesterday but pwired was quicker .

I often use XCloner Standalone as Akeeba Solos didn't already exist at the moment I had to choose for a standalone version, between Akeeba Backup standalone and XCloner standalone.

(I usually have to adjust some "paths" and after that it works well.)

You can exclude chosen directories and files from the backup, and also "Store you backups on the Amazon S3 cloud", for example.

I don't know if Akeeba Solos (which seems to be the new name for the standalone version) can exclude directories and files...

[Peter Knight]

@christophe you can exclude directories and files.

[LostKobrakai]

BittorrentSync could be a rather simple solution for distributing the filesystem to different machines.

[pwired]

BittorrentSync could be a rather simple solution for distributing the filesystem to different machines.

Considering the title of this topic: how would you use bittorrentsync

to backup a database ?

[renobird]

@pwired, I think he was referring to backing up the filesystem by copying it to a remote location. Still on topic.

[Ivan Gretsky]

Was about to try XCloner as I know it back from Joomla times and found this vulnerability report, which seems to be not answered in any way by the XCloner team. Searched exploit-db.com for Akeeba and found just one which seems to be not so important one. Actually I would like to use some free script, but that the way it.is.

By the way there are no entries about "processwire" there.

[Sheri82]

backup through hard disk

[Pete]

I use a variation on some mysqldump commands and zip commands via cron to backup daily to a folder on each site's account, then have a scheduled script on my NAS download the backups each night.

I could do with improving it so it sends reports if it couldn't find a backup file but that's an easy enough thing to check.

I'm sure I could use the programs on my Synology to then backup to the cloud if I wanted to.

Of course if I had hosting at two different webhosts it would all be quicker to transfer backups between them but there's something reassuring about having a backup on a local device

[Peter Knight]

I'm sure I could use the programs on my Synology to then backup to the cloud if I wanted to.

Are you using DSM 5.1?

Synology have a few options for you.

1. Backup to Amazon S3

2. Cloud Sync to G Drive, DropBox, Box, etc

[Skaggs4]

If you want an alternative solution for backup sites and databases then try using CloudBacko Pro.  It works for Windows, Mac, Linux etc and provides some pretty useful features. I am leaving a link with this comment go for it. feel free to use it

Edited January 30, 2015 by kongondo
@Skaggs4: your post looks suspiciously like spam/unauthorised advertisement to me. I have removed the link to the outside resource

[OrganizedFellow]

Revisiting an old thread:

On my Debian box, I have a dump.sh script that I run manually. Also have a git hook pre-commit that runs the same script. Then I git commit to my gitlab account and rsync my site to live site.

I want to automate as much as possible and learn bash as I go.

My next script (once i figure out how) will rsync the latest dump.sql to the live site and import.

---

EDIT, adding below:

On second thought. Since I am using git. Is there a way, to just dump the latest changes in the database, and have that teeny tiny sql file, instead of dumping the whole thing?

Mine is still relatively small, but over time, it will grow.

[Adam Kiss]

I don't think there is anything simple to diff database vs the dump; (additionaly, if you do the diff dump only, you'd then have to diff database versus set of dumps… yeah.)

[OrganizedFellow]

I don't think there is anything simple to diff database vs the dump; (additionaly, if you do the diff dump only, you'd then have to diff database versus set of dumps… yeah.)

http://dbv.vizuina.com/

My Googleism is pretty amazing today, lol.

Looks like a pretty cool project.