Resolved - A network device in Singapore that terminates a portion of BGP sessions experienced TCAM allocation issues. We have implemented a fix.
Our Singapore presence has grown rapidly and as we have now experienced stability issues twice in the same month due to the same device having a resource issue we are bring forward planned works to overhaul our Singapore network presence. This work will now start immediately and includes adding multiple cores in seperate data centers, with a new Equinix SG3 network PoP build already in process, upgraded metro connectivity redundancy, capacity with additional T1 carriers and most importantly removing as many single points of failure as possible.
Sorry for any inconvenience caused.
