BitShares Forum
Main => Stakeholder Proposals => Topic started by: emski on September 12, 2014, 03:44:01 pm
-
I have noticed I have missed blocks for some of the delegates.
There were no network interruptions. I had no ping loses to at least 30 different ip addresses (this is a service running constantly) and this list includes seed nodes and other delegates.
Delegate was connected to the seed node at all times.
Delegate's connections were above 20 at all times.
There were no HDD or CPU spikes that would prevent.
What might have caused missed blocks?
Anyone else experiencing the same?
Furthermore I've noticed some delegates consistently producing blocks with negative latency. Are they manipulating the client in some way? What could've caused this ?
UPDATE: My clock is synchronized but the cliend shows NULL on the time diff. Is this expected?
-
your node can function perfectly and the guy after you can skip you for many reasons including their own connectivity issues, clock issues, etc.
You should not view missed blocks as the fault of the delegate unless they have a streak of missed blocks in a row.
-
your node can function perfectly and the guy after you can skip you for many reasons including their own connectivity issues, clock issues, etc.
You should not view missed blocks as the fault of the delegate unless they have a streak of missed blocks in a row.
Immortal.bitdelegate has 3 missed blocks in a row. Furthermore List_forks shows nothing.
-
I had a strange case too:
https://bitsharestalk.org/index.php?topic=6406.msg113783#msg113783
-
me too
https://bitsharestalk.org/index.php?topic=8507.msg110925#msg110925
-
Odd I just missed 2 blocks in a row apparently, but I've been sitting here watching my client so it wasn't due to lost connection. I've restarted the client and I assume that will fix it but still very strange.
I did leave the client running for a couple days and I know the recommendation was a restart once a day, maybe that has something to do with it?
-
try run:
network_set_advanced_node_parameters {"desired_number_of_connections":100, "maximum_number_of_connections":200}
-
Hmm OK I'll try that out, thanks.
-
Hmm OK I'll try that out, thanks.
Setting desired connection count to 100 is not a good idea. It just puts more strain on your client to service all those connections. I'd personally recommend sticking to 20 or even less (for example, 8 is just fine).
-
Hmm OK I'll try that out, thanks.
Setting desired connection count to 100 is not a good idea. It just puts more strain on your client to service all those connections. I'd personally recommend sticking to 20 or even less (for example, 8 is just fine).
Ok that makes sense, just checked and I have 19 now, usually have between 15 and 20 usually I think.