2

I seem to be having an issue with transactions lately. Here is the most recent one. I paid 0.0004 BTC in fees, and I see there are quite a few inputs, but this seems like a normal transaction and should clear. However, I submitted it through a Multibit wallet about 15 hours ago and it has still not been included in a single block. Two other transactions I attempted recently have also been rejected, not even showing up in blockchain.info. What is the problem?

What, generically, is going wrong here and how can it be prevented?

  • I also have an unconfirmed transaction after 3 hours with 0.0002 BTC fee using MultiBit Classic. – Chloe Jun 15 '16 at 21:48
  • @Chloe: MultiBit Classic development was discontinued in 2015, it only gets security patches. Have you considered updating to MultiBit HD? I'd assume that the outdated software being bad at estimating fees remark I made in my answer would be true here as well. – Murch Jun 16 '16 at 15:18
  • @Murch Duh! I was trying to send my wallet balance to a newer wallet. It kind of sucks that you have to pay a fee just to transfer your own money around. It's kind of scary that one day your wallet could be obsolete and you can't get your money out anymore. – Chloe Jul 4 '16 at 0:54
1

Priority is an outdated concept and most miners don't use it anymore. Similarly, the standard fee is out of use.

Today, miners appear to be selecting transactions based on transaction fee per data size measured in [satoshi/byte].

Your transaction gave a fee of 0.4 mBTC and was 3471 bytes. This computes to 11.5 satoshi/byte. Currently, the predicted fee to get into the next block is 40+ satoshi/byte. Transactions with 11-20 satoshi per byte are predicted to have a 90% chance to be confirmed with a delay of 1-18 blocks.

If your MultiBit wallet is not up to a current version, you might experience better fee estimates by updating. A lot of wallet providers have vastly improved their fee estimations lately.

  • So it's been a fee problem? But for another transaction at 10,439 bytes, I paid a fee of 0.02984783. That's 285 Satoshi/byte, but it still hasn't even been seen by more than 10 peers. Does this seem right? The original transaction finally confirmed, by the way. – 4276 Jun 16 '16 at 14:41
  • @fredsbend: Yeah, that one would have been a fee problem. The new one sounds like a very high fee, perhaps there is another anomaly with it? Would you mind posting a transaction id? – Murch Jun 16 '16 at 15:15
  • Sure, but blockchain and blockexplorer haven't seen it yet: c41da539b77b61ed2f317533d2c0637b565490f143b7c75811b5a7d86dacc97c – 4276 Jun 16 '16 at 16:16
  • I'm on multibit classic 0.5.16. Their site can't seem to tell me if that's old or not. – 4276 Jun 16 '16 at 16:20
  • @fredsbend: It's from July 2015, long before fee estimates were updated. Perhaps it has a non-standard input or something like that? You could submit it directly to blockchain.info to get an estimate. ;) – Murch Jun 16 '16 at 16:30
0

According to the Bitcoin Wiki Transaction Priority is calculated with the following formula:

priority = sum(input_value_in_base_units * input_age)/size_in_bytes

Then you have to appropriately choose your fees based on the resulting priority for your transaction. If you have too low of a priority you should raise your included fees.

Of course, this all depends if the miners are willing to accept your transaction or not. (The sending address could be part of a blacklist, for example. Although I am not aware of such a thing being enforced yet, it could very well be in the future.)

  • So were the fees chosen appropriately in this instance? – 4276 Jun 15 '16 at 22:43
  • 2
    The Wiki is terrible outdated and the priority concept only applies to free transactions, which only a few miners still accept. The only value that really matters for confirmation speed is fee per byte. – Pieter Wuille Jun 15 '16 at 22:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.