Um ex-colega de faculdade e de trabalho era fã de #Python e PostgreSQL.
Minha experiência com #PostgreSQL sempre foi ruim. No começo, eu tinha muita dificuldade em instalar e quando dava certo, não conseguia fazer o #PHP se conectar a base de dados.
Passa o tempo, eu aprendo #Docker e depois #NixOS e decido que a base de dados da minha instância #Nextcloud seria PostgreSQL.
Resultado: toda vez que eu preciso atualizar o sistema, o Nextcloud para de funcionar porque o PostgreSQL não tem um script de atualização próprio. Aí eu passo dois dias ou mais sem poder acessar a minha instância do Nextcloud.
Dessa vez, eu vou abolir de vez o PostgreSQL da minha vida. Só vou usar se for estritamente necessário (ou obrigado) e vou fazer de tudo para não atualizar.
Copiei aqui o 🐚 script que rodo na Ayom, com a principal diferença sendo o uso de contêiner, mas daí também poderíamos adaptar pra usar outro programa. Qual seria ainda mais eficiente que pigz balanceando custo/benefício é a pulga atrás da orelha... :nazare:
Quem quiser permanecer com gzip sem paralelismo pode tentar reduzir o nível de compressão, como no script, para 1 pra ver se melhora o tempo sem perder tanta compressão. Acho que li mais fontes sobre isso, mas uma delas é Peter Eisentraut — Check your pg_dump compression levels.
Both have some minor features, that might be not present in the other, though. E.g. "great invitations", which are ATM only present in Prosody (but WiP for Ejabberd).
currently 1.1G RAM usage for #Pleroma 61G RAM for #PostgreSQL. My instance is 7 years old and running on bare metal hardware. I have tuned #PostgreSQL for performance there.
for multi-user #Pleroma and #Peertube instance #Postgres (v13) uses 3.8G RAM, #PostgreSQL (v15) uses 42M RAM, i need to migrate shit i think. #Pleroma uses 3.8G and #Peertube uses 300M, instance is also on a similar bare metal hardware, but it's much much younger.
@Breadly currently 1.1G RAM usage for #Pleroma 61G RAM for #PostgreSQL. My instance is 7 years old and running on bare metal hardware. I have tuned #PostgreSQL for performance there.
for multi-user #Pleroma and #Peertube instance #Postgres (v13) uses 3.8G RAM, #PostgreSQL (v15) uses 42M RAM, i need to migrate shit i think. #Pleroma uses 3.8G and #Peertube uses 300M, instance is also on a similar bare metal hardware, but it's much much younger.
After supper tonight, I'm going to try upgrading three #PostgreSQL "clusters" from 14 to 15. I don't think there are any anticipated gotchas, but that's why we do these things on little personal databases before taking a wrecking ball to prod.
Still going to have to research how to upgrade a "cluster" that is running streaming replication. That'll be a hassle for sure.
The #PostgreSQL upgrade (14->15) was shockingly easy on all three servers, which makes me wonder what horrors it has in store for me when I go to do the same thing on my work database servers, which have much more complicated access controls, replication, and other fun stuff like that.
Currently I'm working on extending my #Ansible collection more specifically the ansible_podman role.
I'm extending my #podman role with plugins, so it will not only deploy but also configure the containerized application / service.
Currently I'm supporting #postgresql and #borgmatic#borgbackup In that very process I'm also reworking those roles as the plugins just use the existing roles.
Jumpstart a host with postgresql database, borgmatic backup and apprise notifications. #docker
The grist-core repo is the heart of Grist, including the hosted services offered by Grist Labs, an NYC-based company 🇺🇸 and Grist's main developer.
The French government 🇫🇷 organizations ANCT Données et Territoires and DINUM (Direction Interministérielle du Numérique) have also made significant contributions to the codebase.