[BUG] transactions with invalid inputs should never go into mempool #331
Comments
|
all TX from same person: 6733c1ab1e19a9815f00d85c26e90deefe2215f6b0880632e02e3760b4ccc220
12950214c6441709dd2a244211ad353e5d48f0d6736b180444343fe18390ee60
|
|
|
Hello, I have notice this invalid transaction, like: 54f88271901f90ff81fb1d9b75af8a7583d540478e9cad8106ee2b9e71307673 if you cannot open the link, please leave your email address below, I will send by the mail. |
|
can you give a link to your commit for the fix? |
|
diff --git a/src/lib/blockchain/transaction_pool.cpp b/src/lib/blockchain/transaction_pool.cpp
this is the patch |
|
you can ref. the following patch: |
|
step by step build instructions for pool owners: ---> edit src/lib/blockchain/transaction_pool.cpp |
|
Well done again MVS team for not letting pool owners know. Absolutely exemplary work... |
|
it is not mvs teams job to monitor pools. it was a bug and a team member came up with a solution 1 hour after it being reported. metaverse.farm stopped producing blocks almost a complete day before this bug even affected the network, so something else is probably wrong there. Furthermore github is not a good place to whinge and whine, it is mainly used for source code management. However, I understand your failure to realise your pool has not produced a single block in 46 hours and you want to blame someone for it. |
It's mvs teams' job to ensure network stability and conformity is most definitely their job. By allowing serious bugs to propogate, like the one a few months ago where if we hadn't have been notified of a serious wallet bug, that the chain would fork which is pretty serious, they are allowing the network to weaken, thus allowing higher potential of a 51% attack.
I have programmed and used GitHub before you even wrote your first "Hello World" mate. I don't need you to explain to me what it's used for. This is most definitely the place to have the discussion about MVS team's lack of serious bug notifications, the wallets control the network and they are responsible for the network - they have all our email addresses all they need to do is send an email.
Just because you have nothing better to do with your time so you have all this time left to dedicate to ensuring your pool is super operational (might I add while also looking shit and in need of innovation), doesn't mean we all have nothing better to do with our time. |
|
I am befuddled how you managed to not call me "son" in that rant. Everyone in the community is well aware of your tendencies. I mean why on earth would you try to attack me about my first hello world, or "innovation" of dodopool or what I do with my time? I sincerely struggle to understand. I could defend my experience, pool or lifestyle but that'd contribute to this futile argument. Stop attacking people on github or random forum pages. Go innovate some solutions so you don't have to dedicate all your time to pool. on top of my head 3 lines of shellscript i'm not going to get further baited into an argument here, |
Well thanks for confirming I was at least correct you haven't been around long then. The fact of the matter is, regardless of if people LIKE me or not is irrelevant to me and I could care less. What I do care about is a stable, solid wallet. A team which is enterprise and company ready which includes alerting systems for serious network issues. I care about valid transactions, people not ripping off miners like CryptoPoolPond did, or how you somehow used to claim the missing payouts where "out of your control" - was that to pocket some of it? But I don't expect a hobbyist engineer like yourself to know that. |
This was actually confirmed as this bug, as Metaverse.farm started mining again by the patch above... Nothing "wrong" with ANYTHING programmed by me at all. This was after the blocktime attack which I had already notified of and it remained unfixed for some time. Then there was the wallet corruption, and now block validation? The wallet critical bugs have struck again - no enterprise or company would take this seriously at this level of amateur project programming. |
|
that is interesting. can you kindly paste the transaction id that stopped metaverse.farm a day before others? I'd like to see why a transaction only tried to propagate through your wallet and stopped it. |
|
The first time a double spend appeared was I presume the fact the last block was found at 9pm on the 20th was just bad luck that then we were affected by this 12 hours later... (which would be the reason for the length of time between the last block found - 20th at 21:00 - until recently patched....) |
|
See, double spend attempt and this is precisely around the time this all began. All the times add up: https://explorer.mvs.org/tx/6733c1ab1e19a9815f00d85c26e90deefe2215f6b0880632e02e3760b4ccc220 |
|
|
|
After patching It show
|
|
This issue is fixed in v0.8.4. |
Transactions inputs should be validated before going into mempool.
This looks like a critical bug.
At the moment, invalid transaction/input is only detected when that transaction is mined into a block. The network than rejects the block because of the invalid transaction/input naturally. But mining time spent on that block is completely lost.
Example:
.54f88271901f90ff81fb1d9b75af8a7583d540478e9cad8106ee2b9e71307673 goes into mempool at 20180922T000544 and pool starts mining for a solution.
At 20180922T000630 a solution is found but the block is invalid because tx 54f88271901f90ff81fb1d9b75af8a7583d540478e9cad8106ee2b9e71307673 contains invalid inputs. but 46 seconds of mining time wasted to remove this invalid TX from mempool.
The text was updated successfully, but these errors were encountered: