Dgraph debug --postings ./p1/ --perd=City error:Assert failed

Users Dgraph
1

I tested recovering data from v20.11.2 to v21.03.0.
But when I debugged posts for the recovery success time, this error occurred.
What did I do wrong？

root@dgraph2103-dgraph-zero-0:/dgraph/data# dgraph debug --postings ./p1/ --pred=City
Opening DB: ./p1/
Listening for /debug HTTP requests at port: 8080
badger 2021/06/01 08:21:39 INFO: All 2363 tables opened in 237ms
badger 2021/06/01 08:21:39 INFO: Discard stats nextEmptySlot: 0
badger 2021/06/01 08:21:39 INFO: Set nextTxnTs to 4841563
2021/06/01 08:21:39 Assert failed
github.com/dgraph-io/dgraph/x.AssertTrue
     7986cb142993e25a/x/error.go:100
github.com/dgraph-io/dgraph/x.ParseNamespaceBytes
     7986cb142993e25a/x/keys.go:92
github.com/dgraph-io/dgraph/x.PredicatePrefix
     7986cb142993e25a/x/keys.go:506
github.com/dgraph-io/dgraph/dgraph/cmd/debug.printKeys
     7986cb142993e25a/dgraph/cmd/debug/run.go:543
github.com/dgraph-io/dgraph/dgraph/cmd/debug.run
     7986cb142993e25a/dgraph/cmd/debug/run.go:940
github.com/dgraph-io/dgraph/dgraph/cmd/debug.init.0.func1
      7986cb142993e25a/dgraph/cmd/debug/run.go:84
github.com/spf13/cobra.(*Command).execute
        /home/user/go/pkg/mod/github.com/spf13/cobra@v0.0.5/command.go:830
github.com/spf13/cobra.(*Command).ExecuteC
        /home/user/go/pkg/mod/github.com/spf13/cobra@v0.0.5/command.go:914
github.com/spf13/cobra.(*Command).Execute
        /home/user/go/pkg/mod/github.com/spf13/cobra@v0.0.5/command.go:864
github.com/dgraph-io/dgraph/dgraph/cmd.Execute
        7986cb142993e25a/dgraph/cmd/root.go:78
main.main
        7986cb142993e25a/dgraph/main.go:99
runtime.main
        /snap/go/7221/src/runtime/proc.go:225
runtime.goexit
        /snap/go/7221/src/runtime/asm_amd64.s:1371
2

What is a recovery in your terminology? Do you mean exporting all data in v20.11 and uploading it again to v21.03? Or, did you just start v21.03 on the same state at v20.11 (this will not work, which is why I ask).

Also, just inspecting the above snippet, seems to look like the hostname of a kubernetes pod for one of the zero pods, which is not where the data is actually kept, maybe you meant to run this on an alpha pod? (That is just a guess)