User: Password:
Subscribe / Log in / New account

Dirty page caching not mostly harmless

Dirty page caching not mostly harmless

Posted Mar 27, 2014 16:25 UTC (Thu) by zblaxell (subscriber, #26385)
In reply to: Dirty page caching not mostly harmless by dlang
Parent article: PostgreSQL pain points

>> if a process is continuing to make more dirty pages faster than they can be cleaned, such writes can last indefinitely.
> well, that's true no matter what your limit is.

All the more reason to have a plan for this when it happens. Throttle the process dirtying pages so it can't get too far ahead of the block device. Preempt large writes with higher priority I/O (or even equal priority) from other processes.

This can already be done half a dozen ways, but ideally the block layer would default to a more reasonable behavior by itself.

(Log in to post comments)

Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds