Our servers have been getting reset almost weekly, data gets lost, its takes hours to respond to questions. Its impossible to rely on Digital Ocean.
yes its convenient to spawn a server, looks hip. But come on people, its still a technology platform, and we rely on your support to get sh*t done.
Mysql server that has been running for 8 months with no issues, started to fail.
Your customer service guy now suggests not enough ram, so it is shutting itself down.
however the data got lost.. Which does not add up at all.
We have the backups, but it is a very small application, there is no way the data set is too big. There are no writes on this database to increase its memory.
It is annoying to get copy paste responses to your tickets from your technicians.
Are you a consumer product or a professional service?
We obviously understand how mysql and persistence work. it the deamon goes down, data should not be deleted.
DigitalOcean is an unmanaged platform, so you're responsible for whatever you run on your server.
But here in the community forum (user-to-user support), we can try to help figuring out problems, so let's have a go at that.
Are you running using MyISAM or InnoDB for your tables?
If you're using MyISAM, it might lead to corruption if the database crashes, which is a known problem with that type.
If InnoDB, then it's important to configure it, so it saves everything to disk on every query - this makes it a bit slower, but you're safe in case of crash.
Can you post the last 50 lines of the MySQL error log:
tail -50 /var/log/mysql/error.log
Also, you might want to just do a quick check of the database by using MySQLtuner:
https://github.com/major/MySQLTuner-perl
If the Droplet is running low on RAM and MySQL begins requesting more than is currently able to allocated, it is entirely possible for MySQL to crash. When a crash occurs, it is possible that a loss of data can result -- it all depends on the nature of the crash.
I've managed small and large databases over the past 10 years and have ran in to numerous issues -- some that have been fixed in a matter of minutes, some took hours, so I can relate to it being a bit of a burden.
Generally the error log is pretty specific about what's going on and allows us to troubleshoot and prevent these sorts of issues as long as we catch them in time. In such a case, you're already seeing data loss, so I would recommend taking a close look at the error log and if possible, pasting the last 50-60 lines as a reply.
We can then see if there's anything signaling an issue with RAM, or perhaps something more.
tail -50 /var/log/mysql/error.log
Sorry to hear you've been experiencing problems with your servers. Though it's hard to help with out more information. When you say that your server is being reset, do you mean an application you are running is crashing/restarting or that the server itself is being rebooted? Any more details you could provide would make it easier for us to help.