Alternative to disabling timestamp
Posted on 2014-10-30
In a CageScan test, the report advised to disabling "TCP timestamp response" on the F5 LTM but F5 has advised
us against it as given below.
What other alternative we can adopt to mitigate this ? We can't remain status quo as it was not
acceptable to audit.
In particular, can we do anything on Internet facing servers (web servers?), eg, deploy a host/endpoint
based IPS to drop certain pattern or Windows firewall to block certain timestamp traffic or any patch
we can get from MS & Redhat ?
Is this risk considered a DoS vulnerability or ACK storm vulnerability or ?
If we have many servers, is there a quick way to check if "TCP timestamp received"
from ntp server (or is it ntp client?) and the servers' system uptime match?
Can't afford to login individually to each server to check as too many of it & the
uptime data is changing very fast.
F5's response :
"A number of publicly-available networking and security utilities (for example, nmap, hping, and some commercial tools) can perform network scans using a TCP timestamp option, which can be used to calculate the uptime (time since boot) of a system.
Timestamps are a TCP option used by a TCP/IP networking stack to implement two algorithms: the Round-Trip Time Measurement (RTTM) algorithm and the Protection Against Wrapped Sequence Numbers (PAWS) algorithm. Both algorithms are defined in RFC 1323, and are widely implemented by most modern operating systems' TCP/IP stacks, including F5 products.
Because the values in the timestamp are tied to the passage of time as tracked by the system clock, capturing two or more TCP packets allows interpolation back to the time at which the system started.
Uptime information, when combined with other system fingerprinting techniques, may sufficiently identify a system to an attacker as a potentially worthwhile target for an attack.
Note: The issue of using uptime information to select a subsequent attack should not to be confused with any attack against the timestamp mechanism directly.
Eliminating the use of TCP timestamps is not desirable because a performance penalty would occur without RTTM. More importantly, PAWS can protect against both the loss of data when TCP sequence numbers wrap, but also against denial-of-service attacks, which attempt to shut down an existing TCP connection. Without PAWS, the attacker needs only the IP addresses and port numbers of the connection endpoints to reset the connection.
F5 believes that the probability of system uptime information being used as the basis of a subsequent, successful attack is low. Additionally, denying an attacker access to uptime information would be a significant deterrent only if other operating system information could be hidden from fingerprinting techniques, which is not possible. "