New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
What is offloaded MySQL?
Basically a managed MySQL database, you're given credentials and connect and use it without having to manage the underlying infrastructure.
Every shared hosting environment comes with one of these as standard, and various other cloud providers offer them in differing forms.
Any provider offers it but you need to change localhost to the hostname and check if the port is open
Sounds like Amazon RDS.
I think RDS gives each customer their own instance though, so you don't impact other users.
'SharedSQL' is basically just shared hosting in our case. We disable PHP/Apache usually too.
Francisco
More like Amazon Lightsail database. With the usual First 3 month free.
There are use cases for everything I suppose.
For BuyVM shared sql, anyone knows how it is setup? How many databases can we have? Any control panel?
Is it possible to automate entire directadmin server to use BuyVM shared SQL instead of local?
1) It literally runs Directadmin so there's DA + phpmyadmin.
2) this sounds like a good way to get suspended for abuse when a bunch of those accounts spin out of control on disk space/CPU. DA does have an API, but you'd have to integrate it yourself.
Francisco
what is your expected use then; I was considering using it with say a webserver/php on a vm and multiple instances of wp and offlaoding the mysql.
That's fine, people do that.
It becomes an abuse issue if:
Offload SQL was actually a way for us to address performance issues on our OpenVZ nodes. We had a ton of users putting MySQL on their 128MB VM's and then hammering the disk, causing tons of iowait.
SSD's weren't a thing yet, and even when they were it was only 120GB drives in size.
With Offload SQL I was able to throw a lot caching, as well as adding INDEX's to users tables when they gave me permission, to curb a lot of that abuse.
We cut our iowait by over half when we rolled out shared SQL.
These days, our VM's have more RAM, are NVME based, etc, so Shared SQL is more for people that don't want to faff about configuring/managing MySQL and just want a "it just works" solution.
Francisco
And as crazy as it sounds, all this is the reason I have gone back to the insane idea of using flat files (with some secret sauce to avoid too much unnecessary disk io) - which is also a pia for searching.
That's actually RDS with some simpler configs
You can spin up a MySQL CE (or MariaDB) instance on CloudJiffy. I haven't used it enough to know how much extra management it takes over the other options above, but I've used it before when I need a quick RDB. With low activity its just a couple bucks a month.
Until they find a way to make all these so called fancy dbs replicate in different geographical locations and not have a tonne load of lag, all these dbs could go to hell. Sorry, I have been waiting for years.
Just solve for the speed of light, no big deal.
Francisco
Actually I feel there might be something else at play on my opinion but last time I tried it with percona Xtra dB for replication which was years ago, simple ping time differences of about 70ms per location was horrendously slow, I was hoping it was an algorithm thing versus a speed of light challenge. P.S recovery from dbs going offline to resync was a pia so I probably was doing something wrong. Either way, I needed something simple and so I removed the dB overhead for onddns.
Unless your database is on the same lan it is not worth it most of the time because latency hit will be huge and it will slow down everything.
I am already managing my own mysql infrastructure but the connections to it are from LAN. I tried to host a database on public internet and access it, the performance was terrible.
I've used digital oceans shared a.k.a. "managed" database, and it was pretty nice.
it is all about latency, $$ and resources.
If for another $1 per mo you can had off say 5 or 10 dbs with a slight increase in latency it is a good option. But start adding a couple more $ and/or increasing latency then it becomes less attractive.
Only potential downside would be losing the ability to tune mySQL, but everything in life is a compromise.