Running a few days late with the update as I try and shore up some of the issues we've had recently.
The biggest boon for us in Feb was running our our Hive API and relying mostly on that to get honeycomb services going. Our Honeycomb software has been updated to 1.4.3 and is now the preferred way to run DLUX and DUAT(Ragnarok) instances. 1.4.3 includes checking APIs and IPFS at startup and restarts so if your preferred API is down it'll go with the next functioning one on the list. I'm hoping these changes will make a big improvement on overall reliability. These changes were made in response to...
I have been fighting an issue with our IPFS nodes just filling our drives until they drive our servers offline. This really has been one of the most eye opening experiences of IPFS for me. IPFS has been working flawlessly for us for about 4 years now, so the sudden trouble I attributed to all kinds of things I found online. Persistent want lists, maybe the swarm on newer software was trying to make it's own copies. I found out that garbage collection isn't on by default, garbage collection is terribly slow... but even that didn't fix the issue. IPFS pin list CLI is also terrible, with no good way to even see the size of a file and nothing has traditional names, plus terribly documented. But after finding out the our nodes had this big data pinned turned our attention to the install scripts we had been using. Turns out the default config for docker from long ago that I had used at first was a little too permissive, and given the nature of IPFS it's easy enough to crawl the network looking for drive space to store files.
During this time Hetzner decided to ban our IP for using IPFS (it performs a tiny local net scan trying to find IPFS peers on the local network). So off in search of a good VPS/Dedicated solution for our current ecosystem. Found Space Hosting (Affiliate Link) which will even spin up an IPFS on your VPS from the CP... at comparable rates to Hetzner. But this meant moving our many micro-services.
Trying to get DLUX and SPK network back online was a slight pain. Even though our methods for DLUX back-end have long been open-source, a couple of changes on our dApp container haven't been git maintained, so I think I have caddy rules and the dApp container back to the previous security and functionality levels. If you find any issues be sure to let us know.
For the real progress this month, the squashing of several bugs, the improvement of several install scripts, and the writing of more documentation. SPK Network is still very much in test net, but you are all very much invited to take part in the network in anyway you would normally use it. We will try and keep any files/contracts posted on the test net from here on in permanence... which means you should be able to use it as a consumer now. After we switch to live incentives for the main net, those who helped us test and store files on the test net won't be forgotten about.
In the near term the improvement of documentation, and the squashing of bugs is the highest priority I have.
The Storage node includes a Control Panel with a few updated functions. Such as the ability to store all new files in the past 2 days with a click (and sign on keychain).
