cardano-sl node not not syncing new blocks

[tanzeelrana]

/api/settings/sync/progress :

{"Right":{"_spLocalCD":{"getChainDifficulty":{"getBlockCount":544934}},"_spNetworkCD":{"getChainDifficulty":{"getBlockCount":570914}},"_spPeers":0}}

looking at the logs i see :

[node:WARNING:ThreadId 2829503] [2018-02-03 01:57:36.37 UTC] Failed to add headers from "52.193.153.251:3000:0" to block retrieval queue: queue is full

[node:WARNING:ThreadId 2829531] [2018-02-03 01:58:21.11 UTC] Failed to add headers from "52.193.153.251:3000:0" to block retrieval queue: queue is full

[khelle]

I have the same issue. Restart the process helps in my case at least. Alltough I consider it workaround to force restart the node every X hours. I hope devs can give us information how to fix this permamently.

[tanzeelrana]

I had to restart the restart the ./connect-to-mainnet process and it started to sync again as required. I am concerned that

• will this happen again?
• how can we detect it fast enough
• and make sure it will not happen
• why is it happening ?

[gromakovsky]

Which revision did you use? Can you provide more logs (pub folder)?

[tanzeelrana]

AH, I blew up the build and downloaded a latest on today

[khelle]

It happened for me again. It seems this occurs periodcally. Restart helped this time too, but I hope this gets fixed soon. I am using the newest 1.0.3 release. Is it safe to provide the logs of pub folder?

[tanzeelrana]

Restart helped me as well

[tanzeelrana]

could https://github.com/input-output-hk/cardano-sl/issues/2471 be due to this reason ?

[gromakovsky]

Is it safe to provide the logs of pub folder?

Yes, it is.

[top1st]

I have same issue. Some one please help me.

[HM999]

I had this same problem when running on a machine with 2GB of memory. When I increased the memory to 4GB the issue went away. I am posting this in case it helps someone else.