My TestNet BitcoinQT instance appears to be acting a bit weird - it says it is synchronising with the network having over 101000 blocks, while the newest block according to the blockexplorer is 61471. I am not sure if it is some bug in the system, or perhaps just my node is being fed false data.
-
ThePiachu commented at 9:57 PM on March 21, 2013: none
-
Diapolo commented at 10:32 PM on March 21, 2013: none
While you mention it, I observed this 2 times... and thought it was perhaps a one-time glitch or perhaps a malicious node that announced a weird amount of blocks (@core devs is that possible or a problem)?
-
gmaxwell commented at 10:51 PM on March 21, 2013: contributor
It's possible, but it shouldn't be a problem unless we've screwed up: We've been diligent in using that number in only safe ways. As far as I know it's only used in two ways: cosmetically for the progress indicator and, if it's less than the current height, to exclude a peer from preference for pulling the chain.
So at worst someone should be able to do is dork with the progress indicator or cause you to not try to pull the chain from them (which they could have achieved in any case by not responding to attempts to do so).
-
Diapolo commented at 10:55 PM on March 21, 2013: none
The only other place where it is also used, is to display the estimated number of blocks on the network in the debug window. AFAIK we use a median of 10 nodes we are connected to...
-
rebroad commented at 1:17 AM on April 1, 2013: contributor
Some additional checks could be done regarding the current block height. For example, we know a block is generated every 10 minutes, so if they give a number which is higher by n and n*10 minutes is a date significantly into the future, then we can discard this from the median calculation.
-
robbak commented at 1:22 AM on April 1, 2013: contributor
When this happened to me the cause was clients still running one of the earlier testnets. They would report their higher chain length values, and if enough of them connected to you, the number would jump up.
This was considered a useful test mechanism when I brought it up.
On 1 April 2013 11:17, R E Broadley notifications@github.com wrote:
Some additional checks could be done regarding the current block height. For example, we know a block is generated every 10 minutes, so if they give a number which is higher by n and n*10 minutes is a date significantly into the future, then we can discard this from the median calculation.
— Reply to this email directly or view it on GitHubhttps://github.com/bitcoin/bitcoin/issues/2398#issuecomment-15701126 .
-
-
gmaxwell commented at 5:58 AM on April 2, 2013: contributor
We got tired of the "testing" and fixed that. In this case its due to some testing phantomcircuit is doing.
- laanwj closed this on Oct 24, 2013
- MarcoFalke locked this on Sep 8, 2021
