I have understood the problem.
I think that I will use urandom as much as possible.
Thank you all!

> Il giorno 24 mar 2017, alle ore 08:45, Garrett D'Amore <[email protected]> 
> ha scritto:
> 
> Using /dev/urandom to feed entropy to /dev/random is an incredibly bad idea.  
> It may get you moving again, but basically your key material will be far 
> weaker, because the system thinks you’re giving it entropy, when in fact the 
> entropy associated with /dev/urandom is non-existent, except for the entropy 
> that it consumes (from /dev/random!!).
> 
> If you have a microphone, then feeding white noise into it from  /dev/audio 
> recording is better.
> 
> -  Garrett
> 
> On Thu, Mar 23, 2017 at 2:52 PM, ryan nelson <[email protected] 
> <mailto:[email protected]>> wrote:
>  I'll second this recommendation, with a slight tweak:
> Instead of configuring your server/service to *always* use /dev/urandom , you 
> might want to leave it using /dev/random , but "inject" some randomness to 
> get /dev/urandom up-to-speed faster.
> 
> I've had to do this for vm's that were very-recently-launched, to get  
> "openvpn --genkey" to work without a long delay like you've seen with tomcat 
> startup.
> 
> I sped things up with a one-time command of:
> "dd if=/dev/urandom of=/dev/random bs=1024k count=20"
>  ... you can pick your own source of "random enough" data to cat > 
> /dev/random to get its entropy "filled up" quicker
> (with the caveat that you realize what you're doing when you do this, and 
> accept the risk)
> 
> On Thu, Mar 23, 2017 at 9:32 AM, Garrett D'Amore <[email protected] 
> <mailto:[email protected]>> wrote:
>   Actually, the first is a better solution in general.  Even PV devices can 
> generate entropy more slowly; unless you have a very specific need for 
> concrete amounts of raw entropy, /dev/urandom is much to be preferred.  The 
> /dev/random device can block anytime the rate of entropy consumption exceeds 
> the rate of production.
> 
> (However, /dev/random *should* be used if you’re going to seed other pRNG 
> based key generators.  Doing that is generally discouraged if you have a 
> reasonable /dev/urandom implementation though, since /dev/urandom allows for 
> new entropy to be added to the pool over time, whereas other pRNGs generally 
> never increase their entropy.)
> 
> -  Garrett
> 
> On Thu, Mar 23, 2017 at 9:23 AM, Michele Codutti via smartos-discuss 
> <[email protected] 
> <mailto:[email protected]>> wrote:
> Hi all. Recently I noticed that the tomcat web server had log startup times 
> when it runs inside a KVM linux machine.
> It seems that the problem resides in the fact that the /dev/random produce 
> entropy very slowly.
> I have found two solutions of this problem:
>    Configure tomcat to use /dev/urandom
> 2. Use the virtio-rng paravirtual device (if it is implemented in the KVM 
> port in SmartOS). 
> <http://rhelblog.redhat.com/2015/03/09/red-hat-enterprise-linux-virtual-machines-access-to-random-numbers-made-easy/
>  
> <http://rhelblog.redhat.com/2015/03/09/red-hat-enterprise-linux-virtual-machines-access-to-random-numbers-made-easy/>>
> The first solution is quick and dirty.
> The second seems more robust but I need to configure a KVM machine with that 
> paravirtual device.
> I had not found any documentation about that topic in the (SmartOS) wiki. 
> Googling was not useful.
> Some one can give me at least some directions?
> Thanks in advance
> 
> Michele
> 
> smartos-discuss | Archives 
> <https://www.listbox.com/member/archive/184463/=now>  
> <https://www.listbox.com/member/archive/rss/184463/22103350-51080293> | 
> Modify <https://www.listbox.com/member/?&;> Your Subscription

Attachment: smime.p7s
Description: S/MIME cryptographic signature




-------------------------------------------
smartos-discuss
Archives: https://www.listbox.com/member/archive/184463/=now
RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb
Powered by Listbox: http://www.listbox.com

Reply via email to