DaemonForums  

Go Back   DaemonForums > DaemonForums.org > News

News News regarding BSD and related.

Reply
 
Thread Tools Display Modes
  #1   (View Single Post)  
Old 20th December 2014
J65nko J65nko is offline
Administrator
 
Join Date: May 2008
Location: Budel - the Netherlands
Posts: 4,131
Default Attack code exploiting critical bugs in net time sync(NTP) puts servers at risk

From http://arstechnica.com/security/2014...rvers-at-risk/

Quote:
Several critical vulnerabilities in the protocol implementation used to synchronize clock settings over the Internet are putting countless servers at risk of remote hijacks until they install a security patch, an advisory issued by the federal government warned.

The remote-code execution bugs reside in versions of the network time protocol prior to 4.2.8, according to an advisory issued Friday by the Industrial Control Systems Cyber Emergency Response Team. In many cases, the vulnerabilities can be exploited remotely by hackers with only a low level of skill.
For a list of affected operating systems or vendors: http://www.kb.cert.org/vuls/id/852879
__________________
You don't need to be a genius to debug a pf.conf firewall ruleset, you just need the guts to run tcpdump
Reply With Quote
  #2   (View Single Post)  
Old 20th December 2014
jggimi's Avatar
jggimi jggimi is offline
More noise than signal
 
Join Date: May 2008
Location: USA
Posts: 7,983
Default

OpenBSD systems are not affected.
Reply With Quote
  #3   (View Single Post)  
Old 20th December 2014
ocicat ocicat is offline
Administrator
 
Join Date: Apr 2008
Posts: 3,319
Default

Quote:
Originally Posted by jggimi View Post
OpenBSD systems are not affected.
For readers who don't explore the tech@ thread cited, the OpenBSD project includes their own ntpd(8) daemon (OpenNTP) included in base installations. This is different from David Mills' version which is available at http://www.ntp.org/ (which is correctly down...), & which is available as a package/port as net/ntp.

Reply With Quote
  #4   (View Single Post)  
Old 20th December 2014
Oko's Avatar
Oko Oko is offline
Rc.conf Instructor
 
Join Date: May 2008
Location: Kosovo, Serbia
Posts: 1,102
Default

Quote:
Originally Posted by jggimi View Post
OpenBSD systems are not affected.
Reply With Quote
  #5   (View Single Post)  
Old 21st December 2014
rocket357's Avatar
rocket357 rocket357 is offline
Real Name: Jonathon
Wannabe OpenBSD porter
 
Join Date: Jun 2010
Location: 127.0.0.1
Posts: 429
Default

There was a bit of LOC counting that took place, then Theo dropped this gem. I really appreciate what he had to say and I wish more software vendors took the same approach.
__________________
Linux/Network-Security Engineer by Profession. OpenBSD user by choice.
Reply With Quote
  #6   (View Single Post)  
Old 21st December 2014
gpatrick gpatrick is offline
Spam Deminer
 
Join Date: Nov 2009
Posts: 245
Default

Quote:
Apparently the
openntpd math was bad
There may be a reason for the math, but apparently Theo just ignores it. I've also read the code for pf is messy and pf also isn't smp-friendly.

The mantra of OpenBSD is no remote holes but now that some software has been removed from the base (Sendmail, Apache, then ever so briefly, nginx) that had the functionality needed, one has to get it from ports, so the schtick is no longer valid since third-party apps are installed.

I'm not knocking OpenBSD, I use it, but Theo isn't the uber smart, be all, end all of operating systems and security it seems some of you praise him for.
Reply With Quote
  #7   (View Single Post)  
Old 21st December 2014
thirdm thirdm is offline
Spam Deminer
 
Join Date: May 2009
Posts: 248
Default

Quote:
Originally Posted by gpatrick View Post
I'm not knocking OpenBSD, I use it, but Theo isn't the uber smart, be all, end all of operating systems and security it seems some of you praise him for.
That's obviously something of a strawman. It is what it is, but it's a pretty funny coincidence that Theo just happened to take on studying bad pseudo random numbers and seeding and seeing how much could be replaced with arc4random in ports a couple weeks ago (see the tech list archives). Plus they really aren't vulnerable to this one. They also really weren't vulnerable to shellshock.

OpenBSD may not have everything anyone might possibly need or do things always exactly the way every user might wish, but I'm sure wishing I had more time (and skill) to try to figure out how to convert old xaa to newer exa in the old nv driver (or better yet port nouveau over) so I could stop using Slackware on this stupid laptop and get back to OpenBSD. Slackware's nice as linux distros go, but I'm sitting here watching these security vulnerabilities hit the news and finding I'm vulnerable to each one while the system I'd prefer to use isn't.

p.s. on pf not being smp friendly you might also have read or heard that the non-smp OpenBSD pf is faster than FreeBSD smp derivation. SMP isn't the answer to all performance problems.
Reply With Quote
  #8   (View Single Post)  
Old 21st December 2014
gpatrick gpatrick is offline
Spam Deminer
 
Join Date: Nov 2009
Posts: 245
Default

Quote:
you might also have read or heard that the non-smp OpenBSD pf is faster than FreeBSD smp derivation
What you're talking about is here: FreeBSD pf discussion

I'm not going to read each page as I did yesterday, but what you're referring to is in there and there wasn't any data to backup the claim.

I'd say NetBSD's npf is probably the best designed packet filter available today. Mindaugas has thoughts of porting it to FreeBSD and illumos. NetBSD npf documentation

Last edited by gpatrick; 21st December 2014 at 02:59 PM.
Reply With Quote
  #9   (View Single Post)  
Old 21st December 2014
jggimi's Avatar
jggimi jggimi is offline
More noise than signal
 
Join Date: May 2008
Location: USA
Posts: 7,983
Default

Quote:
Originally Posted by gpatrick View Post
.... pf also isn't smp-friendly...
I agree with this. OpenBSD's PF is a component of its network stack, which runs on CPU 0.
Reply With Quote
Old 21st December 2014
ibara ibara is offline
OpenBSD language porter
 
Join Date: Jan 2014
Posts: 783
Default

Quote:
Originally Posted by gpatrick View Post
The mantra of OpenBSD is no remote holes but now that some software has been removed from the base (Sendmail, Apache, then ever so briefly, nginx) that had the functionality needed, one has to get it from ports, so the schtick is no longer valid since third-party apps are installed.
Sendmail was replaced with OpenSMTPD which is really nice if you've never tried it. It's missing a few small things here and there but those things are being actively worked on. Sounds like the OpenSMTPD replacement of sendmail is a great example of how OpenBSD ensures that the mantra isn't a shtick by identifying things in its base system that are not quality and replacing them with things that are.

As to the web server, go find a really good BSD licensed web server that isn't a mess. When you don't find one it seems pretty clear that the only option is to write your own. If you really miss the old apache-1.3, it's in ports as www/apache-httpd-openbsd.

Your statement doesn't hold up to a cursory scrutiny.
Reply With Quote
Old 21st December 2014
rocket357's Avatar
rocket357 rocket357 is offline
Real Name: Jonathon
Wannabe OpenBSD porter
 
Join Date: Jun 2010
Location: 127.0.0.1
Posts: 429
Default

Quote:
Originally Posted by ibara View Post
Sendmail was replaced with OpenSMTPD which is really nice if you've never tried it.
I can +1 this. OpenSMTPD is a great project and has been rock solid for me. I'm excited to see what happens with the in-base httpd as well.
__________________
Linux/Network-Security Engineer by Profession. OpenBSD user by choice.
Reply With Quote
Old 30th December 2014
Carpetsmoker's Avatar
Carpetsmoker Carpetsmoker is offline
Real Name: Martin
Tcpdump Spy
 
Join Date: Apr 2008
Location: Netherlands
Posts: 2,243
Default

Quote:
Originally Posted by jggimi View Post
OpenBSD systems are not affected.
I believe that it is more correct to say that systems running openntpd are not affected; you can run ntpd on OpenBSD, and openntpd on other BSD systems or Linux.
__________________
UNIX was not designed to stop you from doing stupid things, because that would also stop you from doing clever things.
Reply With Quote
Old 30th December 2014
comet--berkeley comet--berkeley is offline
Real Name: Richard
Package Pilot
 
Join Date: Apr 2009
Location: California
Posts: 163
Default

I have never understood why ntpd needs to run all the time in the first place.

In the old days before motherboard clocks with batteries, we used to start up a machine, look at our watch, type in the current date and time and that was it.

These days I automate that process. My /etc/rc.local calls an ntp client to query the time from a time server and then quits.

If I really need the time for a server up 24hours a day, then I can run a daily/hourly cron job to call the ntp client.
__________________
When you see a good move, look for a better one.
--Lasker
Reply With Quote
Old 30th December 2014
Carpetsmoker's Avatar
Carpetsmoker Carpetsmoker is offline
Real Name: Martin
Tcpdump Spy
 
Join Date: Apr 2008
Location: Netherlands
Posts: 2,243
Default

Quote:
Originally Posted by comet--berkeley View Post
If I really need the time for a server up 24hours a day, then I can run a daily/hourly cron job to call the ntp client.
So let's say your clock runs fast, maybe by 3 seconds every day, and then every day your system does a 3-second time travel.
The problem here is that a lot of software relies on reliable timestamps, for example for caching or determining if event A happened before or after event B.

Making you clock jump is *often* okay, but *may* have serious side-effects.

I once had a serious problem in a cluster of servers that did exactly what you suggested (setup by "the previous guy", not me), because a shared NFS drive was used, the mtimes were sometimes slightly incorrect, and one webserver would serve outdated content...
__________________
UNIX was not designed to stop you from doing stupid things, because that would also stop you from doing clever things.
Reply With Quote
Old 30th December 2014
jggimi's Avatar
jggimi jggimi is offline
More noise than signal
 
Join Date: May 2008
Location: USA
Posts: 7,983
Default

There are client/server systems that require timestamps to be closely synchronized. Kerberos comes to mind immediately, I'm sure there are others.
Reply With Quote
Old 31st December 2014
comet--berkeley comet--berkeley is offline
Real Name: Richard
Package Pilot
 
Join Date: Apr 2009
Location: California
Posts: 163
Default

Quote:
Originally Posted by Carpetsmoker View Post
So let's say your clock runs fast, maybe by 3 seconds every day, and then every day your system does a 3-second time travel.
The problem here is that a lot of software relies on reliable timestamps, for example for caching or determining if event A happened before or after event B.

Making you clock jump is *often* okay, but *may* have serious side-effects.

I once had a serious problem in a cluster of servers that did exactly what you suggested (setup by "the previous guy", not me), because a shared NFS drive was used, the mtimes were sometimes slightly incorrect, and one webserver would serve outdated content...
When it comes to assumptions about time, I find it safer to assume that another machine keeps different time from my machine, rather than assume that another machine keeps the exact same time.

For critical systems when synchronizing data from one machine to another, I find it better to use transaction logs and/or version/sequence numbers rather than file modification timestamps (mtime).

In the NFS to webserver example, what happens if a file on the NFS system is intentionally reverted back to an older version of the file from a month ago?

Would the Webserver keep the newer un-reverted version of the file?
__________________
When you see a good move, look for a better one.
--Lasker
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
DoS attacks that took down big game sites abused Web’s time-sync protocol J65nko News 0 9th January 2014 07:34 PM
Security “Bloodsucking leech” puts 100,000 servers at risk of potent attacks J65nko News 0 16th August 2013 07:24 PM
Dual boot with Windows 8: Fast Startup puts data at risk J65nko News 0 14th January 2013 05:31 PM
Security Updates for PostgreSQL 9.1 and 9.2 fix critical bugs J65nko News 0 24th September 2012 04:42 PM
how fast do mirror servers sync to main server for -current ? daemonfowl OpenBSD Installation and Upgrading 3 3rd September 2012 02:58 AM


All times are GMT. The time now is 03:12 AM.


Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Content copyright © 2007-2010, the authors
Daemon image copyright ©1988, Marshall Kirk McKusick