Just got a msg from spacepool, that there are massive problems at the network and my own farmer even has no connections anymore.
Anyone else?
same here,
The newest block on chiastatus is 30 blocks behind my node and all nodes connected to mine
That doesn’t sound good. Strange I just checked, partials are find and blockchain is synced. What version are you on?
(I’m on 1.2.9
I don’t really understand how it all works but I do know that my partials are fine and being updated but block rewards stopped a couple of hours ago. Their home page shows last block won was a couple of hours ago too.
Looks like blocks are still being created on the blockchain, according to my gui at least.
But spacepool didn’t pay out at the usual time. Partials are being accepted though and my unpayed balance is advancing…
Something in the pooling protocol perhaps…
Perhaps it’s just a space pool problem.
no sombody is spaming the whole network with 1 modjo transactions
The network as a whole is fine, just some intermittent issues to increased trasnaction spams of 1 mojo in a world where we don’t yet have self regulating TX fees to curb this.
Here’s a copy of the rundown I posted on Reddit:
I would hesitate to call this an “attack”. Is it annoying? Hell yes. It is problematic? For some individuals. Ultimately though the chain itself is humming along fine and healthy without material impact.
Here’s what’s happening: certain nodes, nodes that are right at or below the minimum spec we established, are struggling to keep up with the 1 mojo spam. A a result they are falling out of sync, and nodes that were already connected to them are struggling to keep up because they are dependent on nodes that can’t keep up. However, the vast majority of nodes are humming along just fine without issue. Furthermore, Nodes that are having no issue on their own keeping up will eventually give up on relying on those weak nodes and drop them in favor of better ones until they correct themselves over time.
This means a select group of people are experiencing pain, but the full chain itself is just fine and unaffected at the core.
This is a short term pain problem in two forms: 1- because the network will self correct, self throttle, and get over this shortly. 2- because this entire situation is only possible because ight now there are no transaction fees.
In the coming weeks this will change. As some stuff we’re working on gets launched, we anticipate a spike in transaction usage. This means we will start seeing transaction fees (however they will be monumentally small at first I expect, fractions of a penny), which makes this spam annoyance improbable for anyone to do more than a few minutes if at all.
Sending out a bajillion individual mojo transactions for free is one thing, but when those transactions cost 10x or 100x or more than that 1 mojo each to send, you suddenly lose the ability to really do this at scale for any meaningful time if at all. Not to mention, legitimate transactions will continue to outbid you, pushing you further down the mempool.
Since right now the blocks are soft capped at 35% capacity intentionally, the need to fees will kick in long before the exceptionally high TX limit of the blocks are reached, by design. Then we can gently raise the capacity as the need raises, keeping fees from snowballing away coughLikeEthGascough and keep things sane for a long time, wile simultaneously letting those fees do what they do well: preventing these annoyances from happening in the future.
4 Likes
all right, I’m going to shutdown my “below minimum spec” node to not slow down everybody and deal with it tomorrow
Thanks for update.
Ppl were spamming btc for kicks at cost for a long time.
Good to stay positive, but saying its improbable is a stretch.
If it means people care? I’ll take short-term pain. If people try and spam then it’s a backhanded compliment. If something doesn’t matter it’s not going to get spammed. Is this not the first plateau along the journey?
But yes, thank you Sargonas for the update. Really appreciate hearing your words on the matter.
Already was wondering why my node stopped submitting partials yesterday… hope it will get back in sync.
What is your time zone?
Thank you for the info. Where can we find those minimum specs?
Also, blaming on general minimum specs, but not directly pointing to which component is sub-spec / lacking right now is not really helping. Most of us run full nodes, where the CPU usage is next to nothing. I guess, that most boxes have at least 16GB RAM (excluding RPis), where potentially just half of that is being used (for dedicated full nodes). Most of home network connections are right now in the range of 100Mbps, and at least for me the biggest spikes I see are in the range of 1-2Mbps. To me, all of those resources are in highly underutilized ranges; however, my ChiaDog is still reporting some issues (maybe along the same time line as @BadgerStork has shown, once we clear up the time zone)
It just sounds like blaming on something just to blame, without any data on hand.
1 Like
I’ve missed 0 challenges, my times haven’t slipped at all.
Its def a hardware issue.
It’ll only become the normal as chia gets more usage.
1 Like
Are you sure that your node stopped submitting those partials? Checking my logs, everything looks good (virtually all partials are processed as usual). Although, checking Flex pool, potentially during the time that @BadgerStork outlined, I have about 60-80% of stalled partials (with the normal number of submitted ones).
Even assuming that there are some issues, when your node will process those partials, my take is that it will pass those results directly to your pool. What it means is that the blockchain side may be out of the picture in this process for the full node, but eventual problems are on the pool side.
Same here. Although, Flex pool sees those partials as stalled (again, all partials are process in time, and internet connection is more or less at idle). I am not sure whether the node reports those partials directly to the pool (that would be my guess), or rather depends on peered nodes (doubt about it).
Again, I would like to have someone point directly to what is lacking. Otherwise, it is just voodoo in place of any real data.
1 Like
That would be my guess also.
Indeed, a list if min specs would help.
I’m currently up to 12mb ram used.
I’d guess for min spec you need
16gb ram
Ssd for chia.
As you said cpu’s normally sit underutilised.
If you meet above spec and still have issues, I guess you then look at ram speed.
I’m swapping my old farmer into a new case today, and adding a ssd, not nice having no fast plotter.
Just hoping she’ll cope with this stress test.
1 Like
My full node a NUC i5-8250U. Not exactly a beast, but also not that shabby. I have 16GB RAM, and usage is at tad over 6GB. SSD is Gigabyte, and usage is again non-existent. Internet is at 100Mbps, usage is either next to nothing or mostly well below 10Mbps (Youtube is putting more traffic, but this is not a fair comparison).
Assuming that RAM speed is slow, that would (slowly ?) cause a backlog, and would end up in debug logs whining about stuff being not handled on time. I don’t see that (RAM runs at 2400MHz - full speed).
One thig that @sargonas mentioned is that part of the problem could be slow peers, and that your client really cannot control, as i doubt that your peers report any such usage to your node.
I am not really whining about my setup (I think it is fine so far, and will not have a problem to beef it up if needed), but rather about having some fuzzy comments about unspecified minimums, without providing any data.
That kind of statements rather bother me. The last part is good (without material impact). However, stating that “the chain itself is humming along fine” where plenty of nodes are struggling is a stretch. I have rather hard time disassociating all nodes from the blockchain itself. If not all those nodes, your blockchain is just useless data siting on an isolated server.
Update
Checking my “Connection” section, and peers that are listed there. It looks like I have plenty of leeching connections. Looking at Task Manager, I have about 10x outgoing traffic than incoming. That looks like a mini DDoS attack to me. My take is that the whole point of peering protocol is that you roughly send as much as you receive. if you are sending more, you have just too many leeches.
You can see up and down speed listed next to each peer.
Oh, you’ve seen that, yep alot seem to limit uploads, maybe internet caps and such.