I have not yet been proved correct in projecting that my maxed-out Slimcoin node will fall behind the chain. It's decidedly busy and seems reluctant to respond to impertinent enquiries about its state of health but it is still synced ... (edited for brevity):
~ $ date; ~/bin/slimcoin getinfo; date
Fri Aug 11 22:35:01 CEST 2017
{
"blocks" : 1070688,
"moneysupply" : 16877871.46784200,
"connections" : 5,
}
Fri Aug 11 22:37:54 CEST 2017
Nearly three minutes to respond to a JSON-RPC command!
~ $ date; ~/bin/slimcoin getinfo; date
Sat Aug 12 10:43:46 CEST 2017
{
"blocks" : 1071200,
"moneysupply" : 16883235.23153300,
"connections" : 5,
}
Sat Aug 12 10:46:16 CEST 2017
Obviously more relaxed at the w/e, two minutes 30s to respond to a JSON-RPC command.
It is still staking like a champ so, to continue the experiment, I'm planning to give it a bit more CPU “oomph” ...
The Slimcoin node is not the only node hosted on our server, it also hosts Minkiz, somewhere around a dozen coin nodes (3 of 'em Slimcoin nodes), *two* instances of ACME (try port 5059: in progress, awaiting me running the scripts to create the RDF graph), an instance of Fuseki, a couple of Abe instances and an Iquidus explorer (that's just off the top of my head).
The i5 that's been doing duty as our server for the last three years is somewhat overdue for replacement. It costs me around 34EUR a month and I can actually upgrade to a faster i7-3770, 32Gb RAM, 2*1Tb HD and save nearly two whole EUR at 32.7EUR a month -
https://www.hetzner.com/sb <- dial the RAM to 32Gb and CPU Benchmark to 9000, choose from there
Thus far it has cost us around 1200 EUR for 3 years of serving web sites, databases, coin nodes, etc. - we wouldn't have been able to fund a bricks'n'mortar outlet, *that's* for sure.
Got some plans for exploiting this edge more intensively, I shall probably start the migration prep over the w/e,
fwiw, I have configured my crontab to cycle over three coins enabling PoW generation for a 10 minute period each (2 per hour). It used to be 4 coins each getting 15 mins but last November, after its upgrade to a masternode codebase, I removed Chaincoin from the setup when the significant increase in diff rendered 15mins of mining inadequate.
(So, why PNut, CHC, FC and RLT? - Pnut was just an overlooked primecoin clone, I wanted to try out prime generation on my lappie and enjoyed building myself up to be a *major* bagholder - which is what happens when there are just 4 nodes in the network. c11 is the same as x11, except that two algos are switched in the ordering, so all the comments about how well x11 does on GPU also apply to c11 and I thought that one day, people might cotton on - that one didn't pan out *quite* as expected but I shan't complain. FC was criticised as a PITA for swapping out *all* the SHA256 for Fugue, not just the PoW hash, and, you never know, it might be the last altcoin standing if SHA256 gets significantly compromised. RLT was horrendously misnamed by a hapless dev and consequently dismissed by the breathlessly avaricious as a lame clone of Litecoin aimed at the gambling industry whereas the roulette reference was to a random choice of algo:
https://github.com/roulettecoin/roulettecoin/blob/master-0.8/src/roulette.h which made it quite CPU-friendly to mine. At some point, I will investigate whether RLT can be nursed along Litecoin's subsequent development path to arrive at a segwit-enabled Bitcoin Core version, same with FC. Ofc, they'll both handle OP_RETURN and so each could have its own click-deployable associated ACME adaptation.)
Cheers
Graham
