toReadableFile: copy error: write |1: broken pipe

[msconnext]

Hello, I encountered an error

2020-12-17T17:06:08.464+0800 WARN ffi-wrapper ffiwrapper/ffiwrapper.go:41 toReadableFile: copy error: write |1: broken pipe
2020-12-17T17:06:08.464+0800 WARN ffi-wrapper ffiwrapper/ffiwrapper.go:51 copied different amount than expected: 65536 != 4161536
2020-12-17T17:06:08.464+0800 WARN ffi-wrapper ffiwrapper/ffiwrapper.go:41 toReadableFile: copy error: write |1: broken pipe
2020-12-17T17:06:08.464+0800 WARN ffi-wrapper ffiwrapper/ffiwrapper.go:41 toReadableFile: copy error: write |1: broken pipe

[msconnext]

Hello,I encountered this error .
My lotus's tag is v1.3.0
this is one of my worker's log when I exxecute lotus-miner sectors pledge
2020-12-17T17:26:59.193+0800 INFO badger v2@v2.2007.2/value.go:1178 Replay took: 1.152µs

2020-12-17T17:26:59.193+0800 INFO main lotus-seal-worker/main.go:335 Opening local storage; connecting to master
2020-12-17T17:26:59.194+0800 INFO main lotus-seal-worker/main.go:392 Setting up control endpoint at 192.168.100.102:2349
2020-12-17T17:26:59.196+0800 INFO main lotus-seal-worker/main.go:495 Making sure no local tasks are running
2020-12-17T17:26:59.242+0800 INFO main lotus-seal-worker/main.go:518 Worker registered successfully, waiting for tasks
2020-12-17T17:27:59.652+0800 DEBUG advmgr sector-storage/worker_local.go:129 acquired sector {{97569 466} 8} (e:0; a:1): {{0 0} /root/.lotusworker/unsealed/s-t097569-466 }
2020-12-17T17:27:59.702 INFO filecoin_proofs::api > generate_piece_commitment:start
2020-12-17T17:27:59.702 INFO filecoin_proofs::api > generate_piece_commitment:finish

2020-12-17T17:28:59.251+0800 WARN ffi-wrapper ffiwrapper/ffiwrapper.go:41 toReadableFile: copy error: write |1: broken pipe
2020-12-17T17:28:59.251+0800 WARN ffi-wrapper ffiwrapper/ffiwrapper.go:51 copied different amount than expected: 65536 != 4161536

[leobelive]

you can check you config ,for you FIL_PROOFS_* config first

[mqqkobe]

@leobelive could you make it?

[0xhhhhh]

$ echo $TMPDIR
/data/tmp
$ mkdir -p /data/tmp

finally restart the worker.

[ribasushi]

This is indirectly fixed by #14