Ranter
Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
Comments
-
zotigapo7646yMigrate your relational db to nosql.
And use cloud services for nosql,
Mongodb
Dynamodb
......
Obvisiously, hosting providers db is on the shared resource (same server).
So if one change effects other customer. -
PaaPsik10416y@helloworld were afraid of amazon because if any fuckups happen and there are too many writes/reads where there shouldn't be - it'll cost an arm and a leg and a kidney. Or am I wrong here?
-
Depends on what you have in that table, what makes row unique and how it is changed you could partition and count rows only on latest partition.
I would also consider increasing timeout. -
did you consider switching to graphQL which doesn't query the whole database every time.
https://www.graphql.com/
Other than that look what AWS, Google or Microsoft has to offer.
https://aws.amazon.com/rds/
https://cloud.google.com/sql/
https://azure.microsoft.com/service... -
@heyheni Pretty sure they are doing that query every morning, because it can have different result and they don't want to get yesterdays results.
-
PaaPsik10416y@Pfannkuchensack I replaced most tables to use inno as it doesn't have table locking
-
Add a Table with the result of the subselect count. So you dont have to count on the big query.
But cant say if this is the right anwser, need some info about the SQL query
Related Rants
PHP dev help/advice needed!
We have problems with mysql. Still stuck with mariaDB, I'm using indexes (correct ones) and we have problems with scaling. we have a few tables with over 100mil rows, 1 of them is being read every morning with a subselect that counts unique rows, and fails every time because of timeout/lock, the temp table size was increased and helped for a little while but as time goes on the table grows and the problem reappears. I'm reading from a slave server that was purposely created for read only, yet we still have problems. We're using managed dedicated servers for out hosting and they aren't willing to optimise the database configs for our needs. What are the easiest options for scaling at this point? Going fully dedicated server and perconaDB? NOsql? Sharding the server? Anyone got any good blogposts or something to read about this? your own experience?
question
nosql
sharding
php
mysql
perconadb
mysql optimisation
mariadb