Julian Elischer
2015-May-05 02:49 UTC
SA-14:19 (Denial of Service in TCP packet processing) and jails issue ?
On 5/5/15 5:28 AM, Mike Tancsa wrote:> On 4/29/2015 6:07 PM, Mike Tancsa wrote: >> >> The IP being scanned is in a jail. If I run the scan to an IP not >> associated with the jail, the scan does not complain. Its only on the >> jailed IP that the scan flags as problematic for this vulnerability. >> >> If this is a false positive, how can I be sure thats the case ? I have >> pcaps of the scan both against the jailed IP (with the scan saying its >> vulnerable) and against an IP not associated with the jail, saying its >> not an issue. >> > > > Anyone have any have any ideas what can be done to mitigate this > risk if its real, or if its a false positive ?Firstly I assume you are not talking about a vimage jail? It seems unlikely that jailing affects that processing. Does the test actually try cause the problem to occur? a tcpdump would be really nice.> > To further clarify/describe my test environment, this is a RELENG_9 > box I am testing against. I have a number of IPs aliased to lo0 > associated with jails. If I run the Qualsys scan against an IP on > this box that is not associated with a jail, it passes the test for > SA-14:19. If I run the test against an IP associated with the jail, > it fails the test. > > e.g. IP 192.168.1.1 is aliased to lo0 and associated with > jail1.sentex.ca. > > If I run the free qualsys scan against jail1.sentex.ca, the test > fails. If I stop the jail, and run the qualsys scan against the > same IP, which is now just an aliased IP on the host machine, it > passes the test. I have the pcaps, but I am not sure exactly what I > am looking for in the data. The test just says it confirmed the > vulnerability with the following 2 tests, > > Tested on port 22 with an injected SYN/RST offset by 16 bytes. > Tested on port 25 with an injected SYN/RST offset by 16 bytes. > > What is it about the jail that might be causing either this issue to > resurface, or give a false positive that its an issue ? > > > ---Mike > > >
Mike Tancsa
2015-May-05 13:32 UTC
SA-14:19 (Denial of Service in TCP packet processing) and jails issue ?
On 5/4/2015 10:49 PM, Julian Elischer wrote:>> Anyone have any have any ideas what can be done to mitigate this risk >> if its real, or if its a false positive ? > Firstly I assume you are not talking about a vimage jail? > > It seems unlikely that jailing affects that processing. Does the test > actually try cause the problem to occur? a tcpdump would be really nice.Hi, Just a plain jail. No vimage. It doesnt make sense to me either how the jail / no jail would impact it. I am guessing some sort of false positive as well, but I dont understand the details enough of how the vulnerability works and why there would even be a different result whether its a jail or not, whether real or not. Here is what I did this AM. In the parent, I stopped the jail and I bound sendmail and sshd to the IP 98.159.241.178 and an instance of apache so the same services would be visible on the scan outside the jail and inside. I then ran the scan. It came up clean. I then removed sendmail, sshd and apache from the IP addresses, and started up the jail # sockstat | grep 98.159.241.178 # /usr/local/etc/rc.d/ezjail start scantest.sentex.ca Configuring jails:. Starting jails: scantest.sentex.ca. 0{vinyl6}# sockstat | grep 98.159.241.178 root sendmail 78661 5 tcp4 98.159.241.178:25 *:* www httpd 78659 3 tcp4 98.159.241.178:80 *:* www httpd 78658 3 tcp4 98.159.241.178:80 *:* www httpd 78657 3 tcp4 98.159.241.178:80 *:* www httpd 78656 3 tcp4 98.159.241.178:80 *:* www httpd 78655 3 tcp4 98.159.241.178:80 *:* root httpd 78651 3 tcp4 98.159.241.178:80 *:* root sshd 78646 4 tcp4 98.159.241.178:22 *:* root syslogd 78586 6 udp4 98.159.241.178:514 *:* # and then restarted the scan. Sure enough, it comes up vulnerable. I have placed the 2 pcaps, and the reports in http://www.tancsa.com/jail ---Mike -- ------------------- Mike Tancsa, tel +1 519 651 3400 Sentex Communications, mike at sentex.net Providing Internet services since 1994 www.sentex.net Cambridge, Ontario Canada http://www.tancsa.com/