What's happening with the Forums

Bykerlass

Well that's as clear as mud to me. Good job there are folk who can sort these things out that's all I can say. Keep up the good work Martin and your little helpers.

mrsyorkie

Keep up the good work Martin - I'd be lost without this site (and even more skint)!

P.S. Southernscouser and Judi - I agree with you both...

joolsybools

Thanks for keeping us posted Martin. I'm happy to offer any help I can, just let me know!

mattlyn73

Thanks Martin and Techies and everyone who contributes to this site for all your hard work.

jamesd

Martin, all versions of MySQL have built in up to the second backup capability using the binary log, so you just don't need hourly backups. See point in time recovery for a general overview of how it all works together.

Daily or so full backups plus retaining the binary logs is sufficient for up to the second backups. Assuming that the binary logs are being written to a different computer, generally via a drive attached via a mount point.

If you don't put the binary logs on a different box you can set the binary log size to a value that causes it to be changed every half an hour or hour, or use FLUSH LOGS to force a log file change, then copy the binary logs to another box with a scheduled cron job. You can do this binary log copying every five minutes if you really want to, since the overhead of FLUSH LOGS is minimal. But it's still second class compared to writing directly to another computer via mounted drive.

Taking hourly backups is only really useful if it takes too long to replay the binary log but since it also requires blocking updates to the database while the backup is being made, if the backup is being made properly, it's almost certainly a bad idea, or not being done properly.

ZTD, have a look at MySQL Enterprise Monitor and the replication monitor. It's a tool available as part of MySQL's support subscriptions, only (not open source). Even without that it's easy enough to have alarms firing if replication stops running or lags excessively. But you're right on the hosting company competence side. It's one advantage of a remote slave, say in the MSE office, where it's easy for MSE people to verify that it's replicating properly as well as delivering an offsite backup source.

Non-techie version: MySQL has built in as standard up to the second backup capability. All you need to do is turn it on and use it. It's worrying that the hosting company even talks about hourly backups when MySQL has up to the second capability as standard.

MSE_Martin

Ta JamesD (and others) I'm making sure my techies who speak the same language are reading

pavlovs_dog

i dont know about anyone else but i've been getting the 'forum closed due to server difficulties' message thingy intermittently for the past hour and a bit

RoCas

Thanks for the info, Martin and a massive well done and thanks to the techies for getting us back up and running.

As a non-technical person who has to deal with our server issues as work, it's quite comforting actually that even the BIG MAN himself has similar problems.

Please remember your health though, Martin, and try not to blow a gasket ... it ain't worth it.

Gilla_2

Thanks for keeping us updated.

We all get withdrawal symptoms when mse is down for long!

May all the nasties who started this live in interesting times!

Alfie_E

Edinburghlass wrote: »

Yep, but I had of course saved the content of the Google cache to my own laptop and have restored the posts today

I had noticed. I’d already said your sterling work deserves a present.