POA: Geth node (miner) automatically stops after 3-4 hours which results in primary node pausing and looking for peers

Question

I have set up 2 Geth nodes with POA on different machines. After starting bootnode and starting Geth nodes, they sync well and starts mining alternately. If I manually stop a node, other node keeps looking for peer. If I restart both nodes, they start working well again. But after 3-4 hours, when they pause miner to " Writing cached state to disk" and "Persisted the clean trie cache", Node2 tries to restart mining and following errors occurs:

0|ethNode2 | ERROR[08-20|10:28:37.169] Etherbase account unavailable locally      err="unknown account"
0|ethNode2 | WARN [08-20|10:28:37.170] Served miner_start                         conn=84.147.159.252:62347 reqid=1134 t="323.102µs" err="signer missing: unknown   account"
0|ethNode2 | INFO [08-20|10:28:37.990] Submitted transaction                      fullhash=0x4336151b175d4d67f68a2760c995ff0c9cb14823c0ba11634cbac1dfb85bbd88 reci  pient=0x352828EF008Cf9f63Cc79ec002449E2d8B16370a
0|ethNode2 | INFO [08-20|10:54:28.648] Looking for peers                          peercount=1 tried=0 static=0
0|ethNode2 | INFO [08-20|10:55:31.673] Looking for peers                          peercount=1 tried=0 static=0
0|ethNode2 | INFO [08-20|10:55:54.716] Writing clean trie cache to disk           path=/home/pistis/ethereum/ethdata/geth/triecache threads=1
0|ethNode2 | INFO [08-20|10:55:54.717] Regenerated local transaction journal      transactions=1 accounts=1
0|ethNode2 | INFO [08-20|10:55:54.747] Persisted the clean trie cache             path=/home/pistis/ethereum/ethdata/geth/triecache elapsed=29.536ms
0|ethNode2 | INFO [08-20|10:55:55.682] Looking for peers                          peercount=1 tried=0 static=0
0|ethNode2 | INFO [08-20|10:56:06.703] Looking for peers                          peercount=1 tried=0 static=0
0|ethNode2 | INFO [08-20|10:56:16.724] Looking for peers                          peercount=1 tried=0 static=0
0|ethNode2 | INFO [08-20|10:56:26.746] Looking for peers                          peercount=1 tried=0 static=0
0|ethNode2 | INFO [08-20|10:56:36.766] Looking for peers                          peercount=1 tried=0 static=0 ``` 


Please note that in the error message `Served miner_start conn=84.147.159.252:62347` This is an unknown IP address so I think the miner tries to connect with an unknown node and is failing to find Etherbase account on that node?

Answer 1

I was able to solve this issue by adding --nodiscover and --netrestrict <CIDR> in Geth Command which helped prevented unknown nodes to connect with my nodes.

For users who still face similar issue even after implementing above suggestion, try reducing RPC-APIs in Geth Command. Only use the ones necessary as there are many bots out there who may try to connect and change your etherbase account, make transactions, etc.

Details here: https://ethereum.stackexchange.com/a/88006/50064