If it seems like web sites have been slow to deal with the Heartbleed bug by revoking and reissuing certificates, turns out there’s a good reason for that: it’s costing them a fortune in bandwidth.
That’s according to CloudFlare, which recently went through the process of reissuing and revoking all of its customers' SSL certificates. In the process of doing so, it got a first-hand look of the effect such a procedure has on bandwidth demands.
As CloudFlare’s Matthew Prince explained in a blog post, the reason for the bandwidth spike has to do with the way a browser checks if an SSL certificate has been revoked.
There are two ways it can be done: Online Certificate Status Protocol (OCSP) or Certificate Revocation List (CRL). The first one simply pings the certificate authority (CA) and asks if the site’s certificate has been revoked. The second downloads a full list of all revoked certificates.
As you might imagine, CRL requires more bandwidth than OCSP. And when you revoke everyone’s SSL certificates en masse, that adds up fast, CloudFlare reports:
Globalsign, who is CloudFlare's primary CA partner, saw their CRL grow to approximately 4.7MB in size from approximately 22KB on Monday. The activity of browsers downloading the Globalsign CRL generated around 40Gbps of net new traffic across the Internet. If you assume that the global average price for bandwidth is around $10/Mbps, just supporting the traffic to deliver the CRL would have added $400,000USD to Globalsign's monthly bandwidth bill. [Bold mine]
CloudFlare ran the same numbers using AWS’s CloudFront price calculator. Using that benchmark, if Globalsign used AWS, the bandwidth costs would have topped $950,000.
CloudFlare also noted that it’s not just a cost issue – a lot of CAs don’t have the infrastructure to handle a bandwidth spike that big.
“So, if you're wondering why some people are dragging their feet on mass certificate revocation,” writes CloudFlare’s Prince, “now you know why — it imposes a real cost.”