Skip to main content

Common signatures

What you seeLikely causeWhat to do
too many open files in the journalfile descriptor limit reachedthe unit sets LimitNOFILE=32768; check open files in status; if the unit was edited by hand, sudo nomctl deploy rewrites it
out of memory, oom-kill, Main process exited, code=killed, status=9/KILL in the kernel journalhost RAM exhausted4 GiB is the minimum; look at MemoryPeak in the bundle's 04-service-properties.txt
no space left on devicedata or backup filesystem fulldf -h; prune backups with nomctl backup --max-backups N, or grow the disk
not enough peers for more than a few minutesfirewallallow 35995/TCP inbound; confirm outbound Internet
synced but the frontier age keeps growingstalled nodesudo nomctl restart; if it recurs, collect a bundle
service running, height not moving, sync state anythingnode stopped syncing momentumssudo nomctl restart; the momentums_stalled alert covers this
leveldb or corrupt errors after an unclean shutdowndamaged chain databasesudo nomctl restore from a backup, or sudo nomctl resync
restarts climbing, nothing obvious in the journalcrash loopsudo nomctl support-bundle --watch and share the bundle
node rpc unreachable ... context deadline exceededRPC accepted the connection but did not answer in 3 secondsthe node is alive but overloaded (heavy initial sync, disk pressure) or wedged; check top for IO pressure, then logs
node rpc unreachable ... connection refusednothing listening on 35997the process is down, or config.json disabled HTTP RPC