On Sun, 2015-08-09 at 12:27 -0700, Bryce Harrington wrote:
Should we implement some of our own additional contingency plans, or are our existing processes already up to snuff?
We have a non-zero contingency, which you saw in action. It's an sqlite db which is in read-only mode and it has issues up the wazzoo; the most significant being the downloads being missing from that db.
I'm thinking we need to get a regular backup going, but recent attempts to dump our database fail with the server killing the process (must be some sort of load protection maybe)
Our staging site could/should operate as a proof for the backups.
I'm thinking like identifying a temporary web host that could serve as an independent cold fall-over in a worst case scenario like this?
Not sure bout web host; that's something beyond these issues with the db. The web server was working well the whole time. It's just our cms and other data is rather db dependent.
Martin,