[Internet]Re: Proposal for potential new feature: TLS Certificate Compression

[Sean Mullan]

On 3/14/22 2:21 PM, xueleifan(XueleiFan) wrote:
> Hi Jamil,
> 
> Thank you for the support of this proposal.  I am happy to move on to 
> the next JEP process if no objections or comments in the coming weeks.
> 
> Hi All,
> 
> Please let me know if more time is required for the review of the proposal.

I suggest you send a new message to security-dev with a link to the JEP 
Draft for review as the next step (even though you already posted it). 
Don't post any code yet, and don't move it to Submitted until there has 
been enough time for review and consensus has been reached that this is 
an idea that is worthy of being added to the JDK Roadmap.

Thanks,
Sean

> Thanks,
> Xuelei
> 
> 
> 
>> On Mar 11, 2022, at 7:25 AM, Jamil Nimeh <jamil.j.nimeh at oracle.com 
>> <mailto:jamil.j.nimeh at oracle.com>> wrote:
>>
>> Hi Xuelei,
>>
>> After reading over your descriptions and the supporting documentation 
>> this certainly seems like it would be a nice addition to JSSE.  It 
>> definitely seems like a performance win for QUIC and also for 
>> bandwidth-constrained devices.  I'm all for seeing this JEP proceed.  
>> I'll likely have more comments down the line and I have taken a look 
>> at the proposed code changes and will go back and do a deeper dive on 
>> it.  I just don't want to jump any of the JEP process hoops by talking 
>> about code changes before some of the important first steps for the 
>> JEP have been cleared.  But you've got my support on this one!
>>
>> --Jamil
>>
>> On 3/7/2022 11:46 AM, xueleifan(XueleiFan) wrote:
>>> Hi,
>>>
>>> The TLS Certificate Compression standard was described in RFC 8879, 
>>> and has been enabled in browser Chrome and Safari. What’s TLS 
>>> Certificate Compression and what’s the benefits of this feature?
>>>
>>> For TLS connections, a client must authenticate the identity of the 
>>> server. This typically involves verification that the identity of the 
>>> server is included in a certificate and that the certificate is 
>>> issued by a trusted entity.
>>>
>>> Where servers provide certificates for authentication, the size of 
>>> the certificate chain can consume a large number of bytes. 
>>> Controlling the size of certificate chains is critical to performance 
>>> and security in QUIC. TLS certificate compression has the potential 
>>> to ameliorate the attacks/problems by reducing the size of the 
>>> handshakes to a size compatible with the security restriction.  The 
>>> TLS Certificate Compression feature is an essential part for QUIC-TLS 
>>> protocols.
>>>
>>> For more details, please refer to section 4.4 in RFC 9001 (Using TLS 
>>> to Secure QUIC):
>>> ---------
>>> Note: Where servers provide certificates for authentication, the size 
>>> of the certificate chain can consume a large number of bytes. 
>>> Controlling the size of certificate chains is critical to performance 
>>> in QUIC as servers are limited to sending 3 bytes for every byte 
>>> received prior to validating the client address; see Section 8.1 of 
>>> [QUIC-TRANSPORT]. The size of a certificate chain can be managed by 
>>> limiting the number of names or extensions; using keys with small 
>>> public key representations, like ECDSA; or by using certificate 
>>> compression [COMPRESS].
>>> ————
>>>
>>> and a more detailed description in the blog “Does the QUIC handshake 
>>> require compression to be 
>>> fast?”(https://www.fastly.com/blog/quic-handshake-tls-compression-certificates-extension-study). 
>>>   I just copy part of the conclusion section of the bog here for your 
>>> quick reference.
>>> ---------
>>> First, the TLS certificate compression extension has a very large 
>>> impact on QUIC performance. Even though the extension is new and 
>>> being introduced fairly late in the process when compared to overall 
>>> QUIC deployment schedules, it seems quite important for both clients 
>>> and servers to implement the new extension so that the QUIC handshake 
>>> can live up to its billing. Without some help, 40% of QUIC full 
>>> handshakes would be no better than TCP, but compression can repair 
>>> most of that issue. I have heard of other non-standardized approaches 
>>> to reducing the size of the certificate chain, and they seem 
>>> reasonable, but this is a problem worth addressing immediately with 
>>> the existing compression extension.
>>> ...
>>> Lastly, data from the real world again proves to be more insightful 
>>> than intuition and is invaluable in making protocol design and 
>>> implementation decisions. When I started this work I expected the 
>>> impact of compression to be positive but marginally focused on a few 
>>> edge cases. The data shows this optimization lands right on the sweet 
>>> spot that ties configurations and the QUIC specification together and 
>>> impacts a large portion of QUIC handshakes. My thanks to the authors 
>>> of the compression extension.
>>> ---------
>>>
>>>
>>> Besides, reducing the amount of information exchanged during a TLS 
>>> handshake to a minimum helps to improve performance in environments, 
>>> for example Internet of Things, where devices are connected to a 
>>> network with a low bandwidth and lossy radio technology.
>>>
>>> This feature is a part to improve the performance of TLS connections, 
>>> and it is also a part of the path towards QUIC standards.
>>>
>>> Chrome support TLS certificate compression with Brotil compression 
>>> algorithm, and Safari support TLS certificate compression with Zlib 
>>> compression algorithm.
>>>
>>> In a summary, JDK could benefits from supporting RFC 8879 in the 
>>> following areas:
>>>
>>>     Performance - Reduce latency and improve performance of TLS and 
>>> QUIC connections by support the TLS certificate compression standard 
>>> in JDK.
>>>     Security - Mitigate the impact of amplification attacks threat by 
>>> reducing the size of the TLS handshakes with compressed certificates.
>>>
>>> What do you think?  Do you want it a part of OpenJDK?  Please feel 
>>> free to share you comments.
>>>
>>> Thanks,
>>> Xuelei
>