For a few weeks now, I've been trying to get my head around what being a witness or running a witness server is really all about. I find myself oscillating between, "Meh, this isn't for me." and "Wow! This is so cool ... I think?"
Yesterday, I realised the only way I'm ever going to understand it -- I'm kinda long in the tooth -- not up with all this newfangled stuff -- is to dive in and try it!
Wish me luck! :-)
I spent three weeks trying.
I used each and every tutorial on Steemit I could find.
I never got a working version up and running which is sad since I own a hosting company and can reproduce the Ubuntu servers when ever I want one.
I even thought about giving someone a free witness server for a year if they would build me one too. That offer still stands but this is the only place I have published it.
Good Luck and if you have not yet started you might want to make a detailed record of your steps.
Again... Good Luck.
Hi @ libertyranger
Yes, some of the instructions are kind of out of date ... and past the time when they cannot be edited. One of my pet hates with forums is how old, outdated information keeps getting indexed by Google AND ends up with higher ranks than the new, up to date stuff, because more visits over time. Grrrr. This does not happen with Wikis! But that's a whole other can of worms.
So, right now I have one witness server and one seed server, both trying to "replay" the downloaded copies of the blockchain. It's taking FOREVER. Three days and counting ... and this is merely replaying a chain they already have fully downloaded. VERY slow.
It seems that the RPC api doesn't come online until the block reindexing is complete ... or at least I hope that's the case. So, I cannot move on to the final steps until that's complete.
The amount of disk space required is a bit crazy, for a hosted server. I have had to split the shared memory file (whatever that is?) and the blockchain data across two volumes, because I've never even seen a web hosting server lease with a large (100G+) boot volume. This seems to be working OK. But it's a significant variation to any of the tutorials thus far.
@privix is hosting servers with suitable configuration out of the box. But having yet another seed server inside the same building doesn't make sense to me. The whole point of decentralization is to, well, decentralize!
Then there's the cost. I'm currently paying US$100/mth, which health/home bound person is more than I can afford. It's clear that I won't be making more than a few SBD a month for a long time. Thus, I doubt I'll be keeping these servers going for long after my initial learning curve.
That said, running a server at home on a 100Mb fibre connection wouldn't be nearly as costly, assuming I already had a spare box to stick it on. But, despite seven computers here, I just don't. :p (They're mostly just too old / slow for this purpose.)
Re the offer, I'm could do what I've done to date for you on a server . for you. However, I think it would be best to get mine fully operational -- to complete the learning experience -- before I made any promises. Even then, it's probably of more value for me to document my process. Give a man a fish and all that. ;-)
Thanks for the reply! Have a great day!
Yes I much prefer the rod and reel method over the gift wrap.
Let me know if nothing else get a witness server at one place and you can put the seed server at my place and we can split whatever the seed server makes and you ride for free on my network
Well, my witness and seed servers finally finished their block replays and started processing blocks, after five days or so. Three months from now, it seems that process will take maybe two weeks ... just to replay local data with 4 CPU cores on twin SSD drives, ffs. A year from now? Two years? Ridiculous!
Anyway, no matter what I do, I cannot get the rpc server to accept connections. Well, more accurately, it accepts a connection but then never responds ...
$ telnet 127.0.0.1 8090 Trying 127.0.0.1... [pour, let cool, sip, then drink two coffees] ^CAt this point, I have to conclude that the software is just lame.
In fact, I'm now at a point where I believe the entire thing is just lame -- almost in scam territory, even. I cannot see Steem having any chance of scaling over the next year (or less) and frankly fully expect the entire network to collapse, once running witness or seed servers becomes simply impractically expensive, if not technically impractical for that matter.
Worse, far as I can fathom after several weeks, it appears that the majority of content creators and people buying SBD are just feeding their coin and efforts to the upper echelons ... under the spell of, "I want to be rich just like them". Ergo, just like any Ponzi scheme. {sigh}
I'll be killing both the witness and seed server after clicking send.
All the best.
Bryan J. Rentoul
Wellington, New Zealand