Livebook crashes after idle period

[euxes]

Environment

• OS: Windows 11 (Chinese Simplified)
• Livebook: v0.19.6, v0.19.7
• Elixir: v1.19.3

Steps to Reproduce

Launch the Livebook app on Windows 11 (CHS). Open a notebook or leave the application idle. Wait for an arbitrary period (crash timing varies). This issue does not occur on Windows 11 (English).

Expected Behavior

Livebook should remain stable and continue running unless explicitly shut down.

Error Log

2026-04-27T10:34:42.062466Z  INFO ** (MatchError) no match of right hand side value:
2026-04-27T10:34:42.062468Z  INFO 
2026-04-27T10:34:42.062469Z  INFO     {:error, {:fread, :integer}}
2026-04-27T10:34:42.062470Z  INFO 
2026-04-27T10:34:42.062472Z  INFO     (os_mon 2.11.1) memsup.erl:843: :memsup.get_memory_usage/1
2026-04-27T10:34:42.062473Z  INFO     (os_mon 2.11.1) memsup.erl:585: anonymous fn/2 in :memsup.handle_info/2
2026-04-27T10:34:42.062476Z  INFO Last message: {:EXIT, #PID<0.1799.0>, {{:badmatch, {:error, {:fread, :integer}}}, [{:memsup, :get_memory_usage, 1, [file: ~c"memsup.erl", line: 843]}, {:memsup, :"-handle_info/2-fun-0-", 2, [file: ~c"memsup.erl", line: 585]}]}}
2026-04-27T10:34:42.062551Z ERROR [os_mon] win32 supervisor port (win32sysinfo): Erlang has closed
2026-04-27T10:34:43.741121Z ERROR Kernel pid terminated (application_controller) ("{application_terminated,os_mon,shutdown}")
2026-04-27T10:34:43.741746Z ERROR 
2026-04-27T10:34:44.031627Z ERROR Crash dump is being written to: erl_crash.dump...done

Livebook.log

erl_crash.zip

[josevalim]

Unfortunately this seems to be an issue with Erlang/OTP. Can you try this:

$ erl
Erlang/OTP 27 [erts-15.2.2] [source] [64-bit] [smp:8:8] [ds:8:8:10] [async-threads:1] [jit]

Eshell V15.2.2 (press Ctrl+G to abort, type help(). for help)
1> application:start(os_mon).
ok
=INFO REPORT==== 28-Apr-2026::19:02:02.607406 ===
    alarm_handler: {set,{{disk_almost_full,"/System/Volumes/Data"},[]}}
2> memsup:get_system_memory_data()
   .
[{system_total_memory,25769803776},
 {free_memory,294174720},
 {total_memory,25769803776},
 {available_memory,9444179968}]

And see if that's enough to reproduce it? Thanks!

[euxes]

Sorry for the late reply.
It seems that os_mon started normally in erl.

PS C:\Users\euxes\AppData\Local\Livebook\rel\vendor\livebook-0.19.7\otp\erts-16.1.1\bin> erl
Erlang/OTP 28 [erts-16.3] [source] [64-bit] [smp:12:12] [ds:12:12:10] [async-threads:1] [jit:ns]

Eshell V16.3 (press Ctrl+G to abort, type help(). for help)
1> application:start(os_mon).
{error,{not_started,sasl}}
2> application:start(sasl).
ok
3> application:start(os_mon).
ok
4> memsup:get_system_memory_data()
   .
[{total_memory,[16949346304](tel:16949346304)},
 {free_memory,[5178179584](tel:5178179584)},
 {system_total_memory,[16949346304](tel:16949346304)}]
5>

Then I had an LLM to write an Erlang script, to keep the os_mon running, and reproduce the issue. Somehow it did catch the exception. I asked if both the crash of Livebook and the exception were tied to OS locale, the LLM gave a yes answer.
Maybe I should report the issue to Erlang/OTP project. 🤣

....
This error occurs when Erlang's io:fread/2 (or similar parser) expects a plain integer but encounters unexpected characters. This almost always happens when the underlying system outputs numbers with locale-specific formatting,
....

[os_mon_monitor] Status -> memsup=true, disksup=true
[os_mon_monitor] Status -> memsup=true, disksup=true
[os_mon_monitor] Status -> memsup=true, disksup=true
[os_mon_monitor] Status -> memsup=true, disksup=true
=ERROR REPORT==== 29-Apr-2026::16:23:03.929000 ===
** Generic server disksup terminating
** Last message in was timeout
** When Server state == [{"OS",{win32,nt}},
                         {"Timeout",1800000},
                         {"Threshold",80},
                         {"DiskData",
                          [{"C:\\",306394108,70},{"D:\\",669442044,55}]}]
** Reason for termination ==
** {{timeout,{gen_server,call,[os_mon_sysinfo,get_disk_info]}},
    [{gen_server,call,2,[{file,"gen_server.erl"},{line,1221}]},
     {disksup,check_disk_space,3,[{file,"disksup.erl"},{line,566}]},
     {disksup,handle_info,2,[{file,"disksup.erl"},{line,340}]},
     {gen_server,try_handle_info,3,[{file,"gen_server.erl"},{line,2434}]},
     {gen_server,handle_msg,3,[{file,"gen_server.erl"},{line,2420}]},
     {proc_lib,init_p_do_apply,3,[{file,"proc_lib.erl"},{line,333}]}]}

=CRASH REPORT==== 29-Apr-2026::16:23:03.929000 ===
  crasher:
    initial call: disksup:init/1
    pid: <0.106.0>
    registered_name: disksup
    exception exit: {timeout,{gen_server,call,[os_mon_sysinfo,get_disk_info]}}
      in function  gen_server:call/2 (gen_server.erl:1221)
      in call from disksup:check_disk_space/3 (disksup.erl:566)
      in call from disksup:handle_info/2 (disksup.erl:340)
      in call from gen_server:try_handle_info/3 (gen_server.erl:2434)
      in call from gen_server:handle_msg/3 (gen_server.erl:2420)
    ancestors: [os_mon_sup,<0.103.0>]
    message_queue_len: 0
    messages: []
    links: [<0.104.0>]
    dictionary: []
    trap_exit: true
    status: running
    heap_size: 6772
    stack_size: 29
    reductions: 8523
  neighbours:

=SUPERVISOR REPORT==== 29-Apr-2026::16:23:03.931000 ===
    supervisor: {local,os_mon_sup}
    errorContext: child_terminated
    reason: {timeout,{gen_server,call,[os_mon_sysinfo,get_disk_info]}}
    offender: [{pid,<0.106.0>},
               {id,disksup},
               {mfargs,{disksup,start_link,[]}},
               {restart_type,permanent},
               {significant,false},
               {shutdown,2000},
               {child_type,worker}]
[os_mon_monitor] Status -> memsup=true, disksup=true
[os_mon_monitor] Status -> memsup=true, disksup=true
[os_mon_monitor] Status -> memsup=true, disksup=true

[josevalim]

Yes, please report it upstream in Erlang/OTP. I would imagine the issue is caused on the path names, since we see these large integers: {"C:\\",306394108,70}.