brooklyn-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Svetoslav Neykov (JIRA)" <j...@apache.org>
Subject [jira] [Created] (BROOKLYN-392) /dev/random fix lost on restart
Date Thu, 17 Nov 2016 09:19:59 GMT
Svetoslav Neykov created BROOKLYN-392:
-----------------------------------------

             Summary: /dev/random fix lost on restart
                 Key: BROOKLYN-392
                 URL: https://issues.apache.org/jira/browse/BROOKLYN-392
             Project: Brooklyn
          Issue Type: Bug
            Reporter: Svetoslav Neykov


In the documentation we recommend users to apply the /dev/random fix. The same is applied
to machines when provisioning.
The problem: the fix doesn't survive a reboot. On reboot device files are recreated. They
don't exist on the file system.

Possible solutions:
  * startup task to do the fix
  * use haveged

Pros and cons:
  * the /dev/random fix always felt hacky to me (Svet). But it works really well. Haven't
seen any security warnings about the approach.
  * haveged seems to be the approach recommended by others. The problem with it is that it
starts burning CPU when there's low entropy. So on a busy system you'd see half of the cpu
taken by Brooklyn, the other half by haveged.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message