Why db file is so huge?  More than 100Mb for some months and very little transacrions.  NXT db now is less than 500Mb  |
|
|
|
Got a major issue with both of my nodes. One is at digital ocean and the other is at home. Today they both stopped syncing all together. I restarted both and after loading the blockchain it restarts and loads it all over again. If I delete the blockchain and have it download it again it loads the very first block then restarts again and remains in this loop. The logs just show the initialization over and over.
I've tried everything, even building a new droplet but the same thing on a fresh install.
I confirm. After deleting blockchain.db client is stuck on genesis block  |
|
|
|
How long to wait Loading ... Please wait!
Blockchain verified complaite, wait synced blocks.? |
|
|
|
Tried to launch Crypti on Cubieboard (with full database from other node)  What to say...  4 hours passed and it still not launched - Blockchain loading: (12668/18072 blocks)  Devs, current implementaition is somewhat very wrong. Not saying about some ecology stuff, but this launching/syncung things are unacceptable  Remember NXT when it went public? It was horrible, there were up to few daily client updates with frequent redownloading of everything. Blockchain was going nuts and resync was almost hourly occurence. It took NXT developers few months to iron all major bugs out but check NXT now - working well (few minor issues) and "glued" to top 5 marketcap. Yes, you are talking with NXT genesis stakeholder.  And i also remember how supportive NXT devs were, how often new client versions come, how they cared about NXT network. Do you see the same here? |
|
|
|
No, but your response confirms my belief
Glad to confirm your fantasies  |
|
|
|
"Iffy thinks i have one, ask him"
"doesn't float your boat" is an expression to mean "doesn't satisfy you?". Much like Беда́ никогда́ не прихо́дит одна́. (a.k.a. When it rains, it pours). For you it's probably best translated as Каждому по своему (to each his own).
Do you need sarcasm tablet? |
|
|
|
Tried to launch Crypti on Cubieboard (with full database from other node) What to say... 4 hours passed and it still not launched - Blockchain loading: (12668/18072 blocks) Devs, current implementaition is somewhat very wrong. Not saying about some ecology stuff, but this launching/syncung things are unacceptable |
|
|
|
sync problems in new ENS code are NOT sync problems in transaction code. I never said that that sync problems were the same. On the contrary, i said straightly what you said - new features had new sync problems. Where is FUD here if you confirm this? Then a new feature was added, ENS. ENS has its own sync channel but is referenced by the transactions model (to ensure you paid for the ENS). ENS is new code, so it is going to have problems....some of which were syncing. Syncing problems in the new ENS code on occasion caused issues in the transaction syncing...eventually they were all fixed. How many new features will be in new beta and how many new "occasion caused issues in the transaction syncing" will be? The more features you have, the more syncing channels you need, and the more possible bugs. I'm saying just the same. Again, is it FUD? every single post anywhere regarding pretty much anything (not just eMunie), you are negative and attempt to twist things to stir up FUD. This is lie. I do not ever recall a single positive or even constructive comment about eMunie...you just bash. Look better, you can find it on eMu forum  Not sure why you mention about having boats.....Mars has no water.
Iffy thinks i have one, ask him |
|
|
|
Prove the recent testing and development have not resolved the sync issues as you claim please. I think it is you who is supposed to give proof in this situation The proof you are requesting will be made available to you and everyone else in just a few weeks. So now there is no such proof for everyone - you confirm my statement. The ability to scale can easily be tested by adding more nodes to the network. Whilst testing we started with a base number of nodes. As the tests scaled up (we added more nodes) the capability of the network was indeed very close to linear. "More nodes" and two more degrees of nodes are some different things, isn't it? Again you do not have to believe me, the proof will be made available to you and everyone else in just a few weeks. I believe not people, but facts. And facts are: almost every new feature added to eMu caused sync proplems in previous betas. And we heard the same claims that "new version will fix it", "the tests in closed circle went good", but new problems arize again and again. Personal attacks? Not from me. If you mistook any of my previous statements as a personal attack I believe you're somewhat over sensitive. Why do on earth you ever began to talk about me? What i have to do, what not, am i greedy, have i a boat, what i should say? |
|
|
|
starik69, please supply current proof to the contrary of what I am saying.
Proof of what? That new beta was not tested on 250 or 2500 nodes? But you alredy know that this is the fact. But in your turn you have no proof of your words - that "ability to scale properly to process transactions at lightning speed" and "will scale almost linear" - this statments are laughable concerning 25 nodes. All yours and others personall attacks on me clearly shows the lack of arguments on your side and my uncomfortable questions stay unanswered Bounties and time period is not worth doing it |
|
|
|
Yes there "was" a problem to start, this has been resolved during testing and development.
There is no proof of that Espesially with all new features Dan coded after previous testing  And history says that there will be syncing problems The use of a block tree means that guaranteeing Consistency (all nodes see the same data at the same time) is not required. So long as a portion of nodes see a portion of the same data at the same time all is good. Different trader swill see different orderbooks? And no syncing problems? Secondly, block tree's allow HIGH Partition tolerance because a block tree is allowed to be partitioned, (or forked in lay terms). HFT trading on partitioned trees? That will be hillarious! Will trades be canselled after resyncing? Or we would be able to doublespend - on one fork and then on another? Why does anyone bother responding to this blatant troll? May be because i am not saying about Dan all that crap you are saying about me? This observation of my style of posting from so honorable eMu adept is noway offtopic in this thread! |
|
|
|
You clearly do not understand the Block Tree design that Dan has explained numerous times over the past year+. For example "fork problems?"; Block Trees are designed to fork, it makes the system far more secure and will lead to a lighter wallet resource wise. Less data to process and store makes syncing less of an issue than you seem to think it is.
Yea tests have never been of any use to indicate real world performance for anything. o.0
Syncing was great problem in previous open betas just after they were "tested" on 25 nodes. This is fact. With all that Trees design. So your argument is irrelevant. Btw, has Dan disprove cap theorem? |
|
|
|
it is virtual machine wit 2Gb and 2 3.40GHz cores After 30 min: Blockchain loading: (5380/16985 blocks) |
|
|
|
But if you not have blocks in 2 hours (or you have 183 blocks), i think something wrong on your node, try to delete your old blockchain and run node again.
183 is after 2 minutes after restart after 2 hours full sync. What with the 0.1.6b link? Why it disappeared? |
|
|
|
Does Crypti comunity has habbit to PGP sing the releases? Can i use db from 1.6 for not to download blockchain again? Edit. Oh, crypti-linux-0.1.6b.zip link disappeared! Whats going on here? |
|
|
|
My node synced for ~2 hours with 1.6 but now all gone - no blocks, no balance. Seems like it is redownloading or reprocessing blockchain from the begining. Waiting 1.6b Btw it is tooo slooow Blockchain loading: (183/16985 blocks)How many hours client will start if blockchain grows to million blocks? |
|
|
|
1. ability to scale properly to process transactions at lightning speed. Test of 1 million transactions in just over 24hrs....done, test complete. We are talking Visa scale capability. As the network grows so does the speed. What is the max you ask.......the network we tested on was 25 nodes in size and spiked to 70 tx/s This will scale almost linear to the number of nodes.
25 is very bad. How you can scale lineary without sync and fork problems? This sandbox tests say nothing about real production |
|
|
|
Yet another IPO from James. How many them already have been? |
|
|
|
Yesterday we had issue with our nodes. Network was work, but not our nodes.
How problems with your nodes is connected to problems with my node? Are you saying that network is not decentralyzed? No, just first nodes to connect written in config file, this is our nodes, later it can be changed to list builded by community nodes, but now network not so large. Maybe you was out of sync because you was connected only to stopped nodes, anyway send me log file via PM, i will check. Restarted and now it looks like synced |
|
|
|
а то переживал что NXT так хорошо развилась в последнее время, а я не купил....
Извините, что со всеми не успел пообщался.
Угу, с мавриками всяко интересней |
|
|
|
|
Show Posts
