{I've set the reply-to to lwig, which I think is appropriate}
SUIT aims at devices where the firmware can be updated as one (or a counted
on fingers few) blob. This is a good constraint, and because it's a "few"
blobs, the edge isn't overly sharp.For instance, we have a common understanding that while SUIT is inappropriate for Smartphone APPs, it is appropriate for the core "System", "Rescue" and "Radio/Broadband" images that are typical for phones. Such smartphones do not fit into RFC7228, and yet they are not "unconstrained" We constrast SUIT to devices where the is potentially many packages that can be updated, up to and including the Linux/Windows desktop/server environment where there are potentially thousands of packages. In RFC7228, we described a series of useful terms and classes, and we have repeatedly come back wishing to have some notions of "class 3+" to describe classes of more capable devices, up to and including "classic" desktop and server OS installations. I think that as we move towards dealing with SBOM concepts (whether via CoSWID, or in liason to IoTSF and/or NTIA) that it would be useful if we worked on an rfc7228bis (or a companion document: nothing wrong with 7228 really), that allowed us to speak more intelligently about different classes of devices. I believe that this should go to the point of having an IANA Registry for the class types, and that RFC8520(MUD) and maybe CoSWID would want to assert such a thing. And probably into some other netmod protocol. Given device FOO on one's Enterprise network, which seems to have a vulnerability, how does one upgrade it? Forklift? JTAG cable? OTA via custom protocol? OTA with SUIT? "apt-get"? "windows-update"? Can device download while it is operational? For instance, my impression is that 90% of Industrial/Smart-City IoT devices in a space way above class 2 (a class 4 or 5!) which are essentially a RPI/Grapeboard/equivalent. In the *best case* running a Yocto build with many many input packages, but only a single image on the output. In a worst case, they are literally Raspberry PI running Raspbian, and dpkg, with the resulting SDcard getting cloned. These devices are hardwired/cabled manually, or experience two-touch onboarding to WiFI or Lora... and they talk back to some cloud provided system that itself may have an unknown set of packages. The SBOM situation could not be worse: it would not surprise me to find gnutls, openssl and gpg crypto on the target system, each with their own copy a RSA and ECDSA encrypt, and should some new oracle/etc. kind of attack to come along, that devices in the field will be completely unevenly patched. -- Michael Richardson <mcr+i...@sandelman.ca>, Sandelman Software Works -= IPv6 IoT consulting =-
signature.asc
Description: PGP signature
_______________________________________________ OPSAWG mailing list OPSAWG@ietf.org https://www.ietf.org/mailman/listinfo/opsawg
