I’m just feeling this out at this point BUT.
I’ve got 38TB allocated to my node presently.
Symmetric uncapped gigabit link.
Over the course of the last few months, I saw traffic spike to a lot early on with the saltlake testing - and once that stopped, now I’m sitting at about 6.2TB used.
I’m getting like 50-100GB of ingress every day and the total usage needle has increased a grand total of… about 50MB (not a typo) in the last THREE MONTHS.
What the actual 
.
If this is the rate at which it’s gonna grow I’m going to be EXTREMELY ACTIVELY looking for something ELSE.
5 Likes
I think you should save yourself the time and disappointment and start looking.
Node growth is variable and generally quite slow.
EDIT: Sorry, just read your post better. 50MB doesn’t seem right. Have you checked for errors on your logs?
Node growth is slow, but not THAT slow!
2 Likes
I guess this is the new normal (for now).
When the tests where running, I added extra nodes to handle the load and to add extra capacity. But sinds the big customer went to Storj Select, I stopped those extra nodes because it make no sense to keep them running. It’s just extra maintenance, electric costs, etc. But eventually everyone has to make their own calculations if it is financially interesting or not.
4 Likes
Yeah honestly I’m starting to feel like Storj Select is basically gonna screw the rest of us…
5 Likes
My current ingress is somewhere around 100GB per day. I started this month with 15.01TB, yesterday it was 15.48TB.
Its not much growth, but more than 50MB per day
2 Likes
Why should it screw us? That doesn’t make any sense. Without select these costumers wouldn’t even onboard here and eventually make ads for us
These are real costumers and not a get rich fast scheme. So if you don’t like it or have a better earn method, than go ahead
3 Likes
A lot of ingress seems to be very short living stuff these days. For this month I have seen 15 TB ingress resulting in 5 TB growth so far.
4 Likes
Yes, but 50MB?
Surely there must be something wrong there…
1 Like
Yeah it seems like data’s getting deleted as fast as it’s coming in or something.
I wonder if I’d get better results if I split this into a bunch of LV’s and ran like ten nodes out of it… pain in the [jerks thumb over shoulder] though.
I can tell you this much with absolute certainty though: the weak point here is neither the bandwidth nor the actual storage stack.
I’m running Ultrastar HC530’s in RAID6 attached to a SAS9116e and cached by a set of SSDs. As my homelab’s network is only 10 gigabits, I can’t even come close to choking that thing out. And the internet link is 1gbps…
Fun fact: I almost crapped myself when I ran the successrate script… and then I remembered I set the logging to “warn” so I’m not logging successes lol (changed that and it looks great again).
1 Like
Hah! That’s actually quite funny!
Any major errors that spring to mind when checking the logs? It’s very odd that you have such performant hardware, are winning races and have such slow growth.
Mind you, I don’t really keep tabs on how fast my nodes grow these days but anecdotally people seem to be seeing a lot more than what you’re reporting… 
Okay yeah THAT I wouldn’t be grumpy to see.
This though:
…WTF.
Sounds like you’re getting considerably more ingress too, what kinda internet link are you on?
You need like 100 /24 subnets to saturate that line… 
1 Like
If you read this forum you surely have read, that everyone has around 50-100Gb a day.
But I don’t know what you mean with usage needle? If you have around 7Tb, how do you know that this is only 50mb increase? (If you mean average disk usage)
Yeah, THAT part is what’s making me wonder WTF I’m doing wrong…
seriously?!
I was thinking about trying it out but… is anyone actually running multiple nodes over for e.g. portmap.io?
My ISP only gives me a single IP (supposedly dynamic but it hasn’t changed in LITERALLY YEARS).
Yes, port map is shit. It has high drop rate and many nodes running on the vpn. And you are breaking ToS
Checking in on it daily for months.
I was at over 6TB usage half a year ago. Currently, 6.28.
So 50-100GB a day is typical ingress?
…is it typical for LITERALLY ALL OF IT to be short-TTL?
So what you’re telling me is…
…to fill this up I’d have to pay my ISP about 25,000 dollars every single month.
These are 2 /24 subnets. So not much better
Yes, if the node is the only one in it’s /24 subnet.
1 Like
