As I described in the thread topic about increasing witness pay, I am committed to increasing my involvement and stepping up my game.
I participated in the recent testnet by building the code, and deploying 2 new fullnodes (8GB, 32GB RAM). I was online in mumble throughout the testing.
I have registered a new domain name for all full nodes (seed nodes, witness nodes and redundant witness nodes) and created a pull request in github to update applications.cpp which contain the seed node definitions.
I have upgraded the node in Australia to an 8GB, quad core CPU to handle the increased trading activity on the DEX. I will also be upgrading another node to 8GB (probably the one in Singapore) later this week.
I have deployed the latest build of the witness_node to all of the nodes.
I began to notice an increase in my missed block count over the last couple of months. Since January 15th 4 blocks have been missed. Reviewing the server load I can see no obvious cause for it. Keep in mind this represents a very low failure rate of only 0.3%. Network latency is reasonable, even for the nodes in Singapore and Australia at between 300 and 400ms.
I am also coming up to speed on Python, and I will be looking into why the latest 0.4.8 version of bts_tools fails to successfully build the witness software from source. It's not a must-have feature of the tools, as the build is pretty easy to do manually now, but it will serve as an aid to getting more proficient with Python as well as contribute a bug fix to wackou's most excellent bts_tools. After that I will turn my attention to python-bitshares. and uptick.