Skip to content

Running full LEMP stack on 128MB RAM, here's what broke

VPS Hosting by Bruno22 79 replies 10K views
#21

Hong Kong to Contabo Singapore is decent latency but the 128MB plan is not worth the transfer. I pay more for local VPS with 1GB and sleep better.

#22

This.

#23

Datenschutz note: Contabo's Singapore datacenter falls under different jurisdiction than EU. If you process EU citizen data, check your DPA. Not a dealbreaker but know where your bits sleep.

Neuland. Aber schnell.
#24

I am in Brazil so LGPD not GDPR but is good point for European members

ping so high I wave back
#25

I test backups by deleting things on purpose and I would never test on 128MB with a database. The OOM killer corrupts things in interesting ways.

#26

SpeedElder here. Old sysadmin. In 2005 we ran MySQL on 256MB servers. But that was MyISAM, no InnoDB buffer pool, no PHP-FPM. Different world.

#27

SpeedElder said:
MyISAM, no InnoDB buffer pool
Exactly. InnoDB's architecture assumes you have RAM to cache. MyISAM was lighter but you paid with table locks and crash recovery. Progress costs memory.

It's always DNS. Always.
#28

One small server, one quiet mind. My quiet mind runs OpenBSD httpd and static sites on 64MB. No PHP, no database, no problems.

rm -rf / --no-preserve-root ☯
#29

zoekyo said:
OpenBSD httpd and static sites
This is the way for 128MB yes. But I want WordPress challenge. Next time I try Ghost or maybe flat file CMS

ping so high I wave back
#30

Ghost on 128MB would also die. Node.js is not lightweight. Try Grav or Kirby if you want flat file.

Post a reply

You need an account to reply. Log in or register to join the conversation.

Post reply Preview Save draft