At any one time when you are connected, you do. Your ISP has a pool of IP addresses which it will dynamically assign to its customers as and when they need a connection. Not everyone needs to be online 24/7. Your ISP will charge you 24/7 for service though and assign you some IP address from their pool as you require connectivity. Its much more profitable that way. Likewise with a mobile connection, you get assigned a public IP “on the fly”.
Great to hear all the tech speak on V6. Can someone who talks tech speak and common English post a few dot points on the Pro’s and Con’s of V6.
I know @Harvindar and I at a minimum would love to know. There must be others.
Very incomplete:
Pros:
-
The world will not run out of IPv6 addresses for a looong time. It is unlikely the Solar System would either.
-
Many networking tasks are much simpler using IPv6 - to the extent where they are not “tasks” at all. Assign the address and get on with it.
-
err thats it
Cons:
-
IPv6 routing - although technically straight forward is not well implemented/supported by anyone other than Chinese ISPs
-
Tunnelling solutions to allow the use of IPv6 addresses across IPv4 infrastructure tend to be buggy and insecure
-
Humans can, at a push, remember and communicate the dotted quads of IPv4 addresses like 154.217.54.23 192.168.1.254 etc. Now try doing the same for IPv6 address like 2a00:23c6:c082:5a00:3b94:b34e:a5d6:6d78
IPv6 has been around for a long time and AFAIK has had std stable implementations for nearly 20 years but has still to gain major traction. I think this is most likely due to the “unfriendliness” of the addresses themselves. Technically, its a good stable solution that I bet most non hardcore networking geeks have at some time said to themselves “Really must get my head around this IPv6 malarkey” and then found better things to do. Sooner or later we will be forced to face up to it.
Thanks for all that you do @sharontannahill
…you rock with the rest of the crew.
![]()
On my side I have this error when trying to join the network. Is this because the network is not accepting more nodes right now, or it’s something else?
[sn_node] INFO 2021-06-18T20:04:58.961028940-04:00 [src/bin/sn_node.rs:112]
Running sn_node v0.49.8
=======================
[sn_node] ERROR 2021-06-18T20:07:58.963183378-04:00 [src/bin/sn_node.rs:143] Encountered a timeout while trying to join the network. Please try again later.
I hope this will help the team.
I never got a node up so I did not get to see the logs but my understanding and to perhaps give you some insight @Cryptoskeptic is that to resolve the out of memory issue from the last test the team have cut down on unnecessary messages and implemented a design where when logs reach a certain size a new log is created, memory is cleared and a new log starts.
From what I read above the system needs some tweaking. Maidsafe please correct me if I am getting this wrong?
You confuse public / private IP and static / dynamic IP.
It does not matter too much if IP is static or dynamic for Safe Network.
Largest testnet file?
safe cat safe://hyfeynycaiekqqtm3orh44aksr9d1p5ekxyniqsexp6f4b8w7rtksjj73ao > 1.iso
Dynamic is most of us right. Can the ISP change it if there is an ongoing connection (node) or can it only change if the connection is lost (does it matter with working IGD)
Hmm. I was wrong. It will make a problems.
But not with initial connection.
ISP will drop all connections every day or so.
If node will be able to recover this event, it will be great.
^ That is for TCP.
What will happen with UDP - I don’t know.
Looks like I understood what are the problems with download timeout ![]()
It triggers not when client is not able to download a chunk, but if total download takes longer than 3 minutes.
But not every file can be downloaded in 3 minutes…
Network do not answers with Retrying after 3 minutes anymore today.
Only Encountered a timeout and unable to establish a connection.
I doubt that it is related to massive upload activity happening during whole night.
Exactly!
And by the way, the explanation of anti-entropy in the OP was a good read. It gave a good overview about the core principles of the network.
Very excited for this release and to see Nikita testing it too!! ![]()
I think it’s died at the mo (or really limping along). The log rotation didn’t make it to the droplets. So could be that having slowed down things. Genesis appears to be quite out of sync with expected balances atm too.
Some improvements inbound. Not sure if we’ll get them up the weekend though
time safe cat safe://hygoyeycjibzowe4bizoherbc9kn56ojde9skr7xu3hw5u399pnc8eeiyhw > Safe_Put_Cat/Safe_Nugget_catted.png
this command still work!
Wow that was FAST!
time safe cat safe://hygoyeycjibzowe4bizoherbc9kn56ojde9skr7xu3hw5u399pnc8eeiyhw > Safe_Nugget_catted.png
real 0m6,684s
user 0m0,360s
sys 0m0,166s
Yes @Josh , these are a couple of things
. There’s been (and vigourosly continues to be) scrutiny all over the place, high and low, to cut down on redundancies, complexity and wastefulness. Many of these fixes/changes will have a positive impact on performance, some will make the code simpler, less error prone and easier to maintain. It goes hand in hand often.
@joshuef @oetyng A lot of quick fixes now and code clean up. AE happening properly now and @bochaco is massively simplifying the whole messaging story (never made testnet but we are checking sigs twice per message → mental). A huge improvement incoming is the mono crate Josh pushed. That puts routing/node/client/messages/data types etc. in a single repo. This means no more version inconsistencies plus the biggie, any changes to any part will force an end to end full network test.
With that, we will have fewer “silly” bugs (I say silly as these are the bugs we see in these testnets and
) and have many more specific and exact bugs/tweaks we can try like section size and so on. It’s really great to be in this place and with new team members bolstering the ranks it feels like we have executed this part of the plan really well. We held the thin blue line, now the battle looks well and truly in our favour. With a small dedicated capable team this really recovers my faith in humanity. There will be mega
soon ![]()
safe keys create --test-coins --for-cli --preload 1000000
