otto

joined 2 years ago
 

What are your preferred strategies when a MySQL/MariaDB database server grows to have too much traffic for a single host to handle, i.e. scaling CPU/RAM or using regular replication is not an option anymore? Do you deploy ProxySQL to start splitting the traffic according to some rule to two different hosts?

Has anyone migrated to TiDB? In that case, what was the strategy to detect if the SQL your app uses is fully compatible with TiDB?

[–] otto 3 points 2 days ago (1 children)

By UV 3000 you probably don't mean the ultraviolet lamp that is the first page of Google is full of when searching with this term..? I doubt UV - whatever it is - is a common approach.

 

What are your strategies when a MySQL/MariaDB database server grows to have too much traffic for a single host to handle, i.e. scaling CPU/RAM is not an option anymore? Do you deploy ProxySQL to start splitting the traffic according to some rule to two different hosts? What would the rule be, and how would you split the data? Has anyone migrated to TiDB? In that case, what was the strategy to detect if the SQL your app uses is fully compatible with TiDB?

[–] otto 2 points 2 days ago (1 children)

What do you mean a default user? You can just run 'mariadb' to access to console with the same user that had permissions to run 'apt install'.

For your actual application you need to plan what database name to use, what user, what permissions it needs, potentially remote connection and TLS etc. This indeed is some work and could perhaps be automated a bit, but it also needs sysadmin to make some decisions.

 

Besides having the latest version available, what do Debian users who run MariaDB wish to see in future versions of MariaDB, or how it is integrated and packaged in Debian?

I am the maintainer in Debian - looking for feedback and ideas.

[–] otto 2 points 2 days ago (1 children)

Yes, increasing the InnoDB buffer pool to use all available memory is the most important configuration change a sysadmin can do. But in order to do it, you need to know if the host is dedicated to one MariaDB instance or if there are multiple servers on the same host. Otherwise you would just have processes each hogging more memory when they can and not giving it up to others.

I could think about having a dialog during the installation that asks something like "Is host dedicated to this MariaDB instance? If yes, automatically configure it to use most of the system RAM available."

[–] otto 1 points 2 days ago

MariaDB supports Galera clustering out-of-the-box, and also traditional primary/replica setups. But you need to have something that spans multiple hosts to monitor and manage it, and that is outside of what a single-host OS package management system can do.

 

Besides having the latest version available, what do Debian/Ubuntu users who run MariaDB wish to see in future versions of MariaDB, or how it is integrated and packaged in Debian?

I am the maintainer in Debian and Ubuntu - looking for feedback and ideas.

[–] otto 1 points 3 days ago (1 children)

You mean ollama? There are so many options, any favorites?

 

Besides having the latest version available, what do Ubuntu users who run MariaDB wish to see in future versions of MariaDB, or how it is integrated and packaged in Ubuntu?

I am the maintainer in Ubuntu - looking for feedback and ideas.

 

Besides having the latest version available, what do Ubuntu users who run MariaDB wish to see in future versions of MariaDB, or how it is integrated and packaged in Ubuntu?

I am the maintainer in Ubuntu - looking for feedback and ideas.

 

What are your strategies when a MySQL/MariaDB database server grows to have too much traffic for a single host to handle, i.e. scaling CPU/RAM is not an option anymore? Do you deploy ProxySQL to start splitting the traffic according to some rule to two different hosts? What would the rule be, and how would you split the data? Has anyone migrated to TiDB? In that case, what was the strategy to detect if the SQL your app uses is fully compatible with TiDB?

 

I’ve been exploring MariaDB 11.8’s new vector search capabilities for building AI-driven applications, particularly with local LLMs for retrieval-augmented generation (RAG) of fully private data that never leaves the computer. I’m curious about how others in the community are leveraging these features in their projects.

I’m especially interested in using it with local LLMs (like Llama or Mistral) to keep data on-premise and avoid cloud-based API costs or security concerns.

Does anyone have experiences to share, in particular what LLMs are you using when generating embeddings to store in MariaDB?

 

Another release to round out this month. Enjoy.

 
 

The XZ Utils backdoor, discovered last week, and the Heartbleed security vulnerability ten years ago, share the same ultimate root cause. Both of them, and in fact all critical infrastructure open source projects, should be fixed with the same solution: ensure baseline funding for proper open source maintenance.

 

The XZ Utils backdoor, discovered last week, and the Heartbleed security vulnerability ten years ago, share the same ultimate root cause. Both of them, and in fact all critical infrastructure open source projects, should be fixed with the same solution: ensure baseline funding for proper open source maintenance.

[–] otto 4 points 1 year ago

I just prefix all my git aliases with g-. So for status I type g-s<tab>.

[–] otto 2 points 1 year ago

You need bisect only as a last resort. Effective use of git blame, git log -p -S <keyword> etc has always been enough for me. Also, the projects I work with take 10+ minutes to compile even when cached, so doing tens of builds to bisect is much slower than just hunting for strings in git commits and code.

[–] otto 4 points 1 year ago

I had the same feeling until I started using gitk. I always have a gitk window open and press F5 to reload, so it shows me the state of everything after I've run git commands. Now I grasp everything much better.

[–] otto 1 points 1 year ago

Only product from Microsoft I actually like using and trust. Quality from 1998, and still going :)

[–] otto -1 points 1 year ago

One is enough if it is very big

[–] otto 1 points 1 year ago

Try again tomorrow, seems it got popular today

view more: next ›