[systemd-devel] Not start readahead-collect if /.readahead exists

[Lennart Poettering]

On Thu, 14.04.11 17:35, fykcee1 at gmail.com (fykcee1 at gmail.com) wrote:

> Hi all,
> 
> /.readahead exists means activated readahead-replay will run. Meanwhile if
> readahead-collect runs, all replayed blocks will be collected -- size of
> /.readahead will increase(e.g. in case of services update).

Well, actually this was intended to work the way it currently does:
i.e. run the collector and replay at every single boot. While this of
course will only increase the readahead size over time this should
behave asymptotically.

To make this more robust we might want to add a different fix here. For
example, we could use certain important OS files as flag files to --
when they change ctime -- drop the readahead file for the next reboot
(RPM database?). Or we could compare /etc/os-release with the previous
one. Or something like that. For simple individual package upgrades the
readahead data shouldn't really outdate so quickly, and readaheading a
little bit too much should be cheaper than a little bit too little.

Lennart

-- 
Lennart Poettering - Red Hat, Inc.