JRA Website Projects
JRA & Members' Projects
Other Projects & Events
Photos - Junks Old and New
Public Domain Files by .....
Design & Navigation Software
Members' Blogs & Websites; Tracking Systems
Sails, Masts & Gear
Weather, Charts, Stars & Guides
Sailplans - Current
Sailplans - Early
HALL OF FAME
2010 UK East Coast
2010 UK South Coast
2010 UK Wales
SWAP, SELL or BUY
Back to topics
Site Backups & Security
Show latest replies on top
Subscribe to topic
03 Feb 2014 21:46
From time to time I'll be updating this thread with info on how JRA volunteers backup the amazing and in some cases irreplaceable range of data, fora posts, magazines, images, links, etc that we've built up on the site since it's inception.
The Entire Site
Each page is or is represented by a file which is hosted on the Canadian company Wild Apricot's servers. Those charged with updating site pages are advised to create a backup copy of each page they update, before updating it, or of each new page. Those copies are also hosted on the site, since it is not straightforward to download the entire contents locally. The idea is that we should always have a page to revert to.
"We automatically back up all Wild Apricot accounts on a daily basis and store the backups at a location separate from our data center. These backups protect your data from catastrophic threats to our data center, and can be used to restore lost data only in the case of emergency. These backups are not available to clients."
Despite that last statement, in emails they have said that they
can perform a next-day restore of any page or file that we mess up or
lose. As far as I know we have never needed to test that service.
WA recommend a third-party full site contents backup service. Paul Thompson and Roy Denton and I have experimented with that. While it is useful for taking an 'image' of the site's contents, it is not straightforward to use and, more importantly, does not create a resource that we could use to rebuild the entire site and it's structure: at the moment only WA can do that. When Paul has the time he hopes to experiment with the difficult job of writing some software that will create a full-restore copy. Why? We are dealing with a commercial supplier here -WA - and while we have no reason to doubt their viability, any company can fail, and government or private hackers can cause damage.
(Wild Apricot say that their main data center runs on an enterprise hosting platform by CBeyond (formerly MaximumASP) in Louisville, Kentucky. You can find additional information regarding the security on our page here
Backups of our magazines from #1 to the current edition are archived by Paul Thompson in New Zealand and Roy Denton in the UK, so we have a set on each side of the planet. They are hosted for site download purposes on www.box.com, where we also store the originals of issues created by editors Lynda Chiddell and David Tyler in Apple's 'Page' format. David and Lynda no doubt have copies on their own machines too. This means that we are able to recover corrupted magazine files - it happens - within a day, and often within an hour or two.
Images and Documents
These are currently stored on the site, with backups (perhaps) available on demand from the computers of those who provided them and/or who uploaded them to the site. Such files may be caught by the third-party backup software referred to above. Clearly there is room for improvement here, though we have of course WA's assurance that these can be restored from their previous day's backup of the entire site.
All Committee members have full access to the site and could edit or help edit any page on it should they so wish. Only David, Lesley and I ever do that. Do we want to retain full admin status for all? The more admins we have, the more potential loopholes there could be.
(more to come)
Last modified: 05 Feb 2014 23:22 |
Back to top
" ...there is nothing - absolutely nothing - half so much worth doing as simply messing about in junk-rigged boats"
the Chinese Water Rat
Site contents ©