Zudoku
Cloud Platform

Troubleshooting

This guide explains how to debug apps on Unikraft Cloud so you can fix issues or collect enough information for the support team to help you. It follows platform best practices and supports efficient troubleshooting. If you need help, reach out to Unikraft Cloud Support.

Debugging a problematic app

An app may crash, freeze, or misbehave. To inspect it, try either:

(bash)
 
kraft cloud inst logs <instance_name>
# or
kraft cloud inst get <instance_name>

This may, in certain times, provide insufficient information.

Large filesystem build process gets stuck

When the filesystem is larger than about 800 MB, the build may get stuck. A limitation in BuildKit, the current filesystem build component, causes the issue.

Work around this by reducing the filesystem image below 800 MB. Work is in progress to integrate a component without this limitation.

Gateway error when accessing a service

When you query a Unikraft Cloud service via its public address, such as:

(bash)
 
curl https://frosty-bobo-zeev783o.fra.unikraft.app

you may get a Bad Gateway response:

(html)
 
<html>
<head><title>502 Bad Gateway</title></head>
<body>
<center><h1>502 Bad Gateway</h1></center>
<hr><center>openresty</center>
</body>
</html>

This happens when you specify the wrong internal app port (for example, the app exposes port 8080 but you use 443:80 instead of 443:8080).

Another error response you may get is:

(html)
 
<!doctype html>
<html lang="en">
<head>
        <meta charset="UTF-8">
        <meta name="viewport"
              content="width=device-width, user-scalable=no, initial-scale=1.0, maximum-scale=1.0, minimum-scale=1.0">
        <meta http-equiv="X-UA-Compatible" content="ie=edge">
        <title>Service not found</title>
</head>
<body>
There is no service on this URL.
</body>
</html>

This happens when you connect with an HTTPS client (port 443) and the app doesn't expose that port. Database services such as MongoDB or MariaDB use different ports (for example, 27017, 3306).

Use the correct exposed port. You may need a TLS tunnel (see below).

Connect to a non-TLS app

Unikraft Cloud uses TLS to expose services to the outside world. Some apps (such as MongoDB or MariaDB) don't use TLS. Create a TLS tunnel via kraft cloud tunnel, which opens a local endpoint (localhost / 127.0.0.1) and forwards traffic over TLS. See the MariaDB guide for an example.

"No such file or directory" when building or deploying an image

When building / deploying an image, you may get the error below:

(ansi)
 
 E  unpacking the image: opening layer: open <path_to_home_directory>/.local/share/kraftkit/runtime/oci/digests/sha256/3d30a7ba2a819aec2e73c5df07c24264d66891e926b395b9ef0f66f151db4b49: no such file or directory

This often means the local kraft cache is in an inconsistent state.

To solve this, remove the local kraft cache and local packages:

(bash)
 
kraft pkg rm --all
rm -fr ~/.local/share/kraftkit

Launched app not visible in list

The most common reason is that you deployed an app to one metro but listed a different one. Set the metro for a session with an env variable:

(bash)
 
export UKC_METRO=<metro_name>

or for each indvidual command via the --metro flag, for example:

(bash)
 
kraft cloud instance list --metro <metro_name>

How can you cache the app's filesystem for faster builds

When using a Dockerfile for the app filesystem, Unikraft Cloud passes the commands to BuildKit.

By default, each kraft cloud deploy command starts an ephemeral BuildKit container. The platform then removes the filesystem app data, so each deploy starts from zero.

To prevent this, follow the instructions here:

(bash)
 
docker run -d --name buildkitd --privileged moby/buildkit:latest
export KRAFTKIT_BUILDKIT_HOST=docker-container://buildkitd

The approach above caches builds in the BuildKit container filesystem. Another approach saves the cache in a local host directory:

(bash)
 
docker run -d --name buildkitd --privileged -v $HOME/.buildkit-cache:/var/lib/buildkit moby/buildkit:latest
export KRAFTKIT_BUILDKIT_HOST=docker-container://buildkitd

Here $HOME/.buildkit-cache is a local path on your machine where BuildKit stores cache data.

What's a Kraftfile

A Kraftfile is used by the kraft CLI tool to understand how to build and deploy your instance. Typically you can use the default Kraftfile found in each Unikraft Cloud example. Below is a sample Kraftfile with a brief explanation:

Kraftfile(yaml)
 
spec: v0.6

runtime: python:3.12

rootfs: ./Dockerfile

cmd: ["/usr/bin/python3", "/src/server.py"]

labels:
  cloud.unikraft.v1.instances/scale_to_zero.policy: "on"
  cloud.unikraft.v1.instances/scale_to_zero.stateful: "true"
  cloud.unikraft.v1.instances/scale_to_zero.cooldown_time_ms: 1000

The runtime specifies one of the Unikraft Cloud runtimes (unikernels) built to run different languages and apps. Here it specifies a Python runtime.

The rootfs parameter tells kraft to use a Dockerfile in the same directory to build the root filesystem, and the cmd parameter which command to run when deployed (although you can also specify this in the Dockerfile).

Finally the labels specify runtime options. In the example above, the configuration enables stateful scale-to-zero and the platform waits 1 second with no requests before putting the instance to sleep.

Debugging the build, packaging and pushing steps

The kraft cloud deploy command performs steps on your local device before actually deploying your app to Unikraft Cloud:

  1. Downloads the runtime image (defined in the Kraftfile) from the Unikraft Cloud registry.
  2. Builds the app filesystem using the Dockerfile via BuildKit.
  3. Packages the app filesystem and the runtime in an Open Container Initiative (OCI) image.
  4. Pushes the OCI image to the to the Unikraft Cloud registry under your username's namespace.

If any of these steps fail, enable kraft debugging with the --log-level and --log-type flags:

(bash)
 
kraft cloud deploy --log-level debug --log-type basic [...]

You should then see debug output for the 4 steps above.

(ansi)
 
 D  using metro=fra
 D  using token=***
 D  cannot run because: no image specified deployer=image-name
 D  using deployer=kraftfile-runtime
 D  querying catalog local=true name=index.unikraft.io/razvan.unikraft.io/http-go-strace2:latest remote=false
 D  found 0/66 matching packages in oci catalog
 D  using packager=kraftfile-runtime
 D  querying catalog arch=x86_64 local=true name=index.unikraft.io/official/base plat=kraftcloud remote=false version=latest
 D  found 0/66 matching packages in oci catalog
 D  querying catalog arch=x86_64 local=true name=index.unikraft.io/official/base plat=kraftcloud remote=true version=latest
 i  found index.unikraft.io/official/base:latest-dbg (kraftcloud/x86_64) (6bd4542, 11c196e, 2.0 MB)
 D  found 1/67 matching packages in oci catalog
 D  saving manifest digest=sha256:6bd45428b8ecf667954cb8f829ee3f8f6f6f23fa0a0e7dc1aa27fcae39826f89 ref=index.unikraft.io/official/base:latest
 i  building rootfs
 D  using buildkit addr=docker-container://buildkitd version=v0.12.3

This output should give enough context to diagnose the issue. If not, report it with the output on the Discord server.

Debugging running apps

The most direct way to debug an app is to use the app console output, which may include kernel output. To see it, after starting the Unikraft Cloud instance (via kraft cloud deploy or kraft cloud inst create), use:

(bash)
 
kraft cloud inst logs <instance_name>

In case of a crash, you'll see a full crash output:

(text)
 
Powered by Unikraft Telesto (0.16.2~9c264902)
[    0.066949] CRIT: [libukvmem] Cannot handle write page fault at 0x1000bb8024 (ec: 0x2): -12
[    0.067680] CRIT: [libkvmplat] page fault handler returned error: -12
[    0.068227] CRIT: [libkvmplat] Unhandled page fault vaddr=0x1000bb8024, error code=0x2
[    0.068910] CRIT: [appelfloader] Unikraft crash - Telesto (0.16.2~9c264902)
[    0.069503] CRIT: [appelfloader] Thread "python3"@0x4017e4020
[    0.069993] CRIT: [appelfloader] RIP: 0008:000000100041fdcb
[    0.070468] CRIT: [appelfloader] RSP: 0010:000000100017f760 EFLAGS: 00010206 ORIG_RAX: 0000000000000002
[    0.071269] CRIT: [appelfloader] RAX: 0000001000bb8000 RBX: 00000004018012f0 RCX:0000000000000007
[    0.072040] CRIT: [appelfloader] RDX: 0000000000000000 RSI: aaaaaaaaaaaaaaab RDI:00000010007bccc0
[    0.072793] CRIT: [appelfloader] RBP: 000000000000001a R08: 0000000000000001 R09:00000010004e2b99
[    0.073558] CRIT: [appelfloader] R10: bb9e744ce6503a27 R11: 0000000401861290 R12:00000010007bccc0
[    0.074312] CRIT: [appelfloader] R13: 00000010007bbd48 R14: 000000100017f890 R15:0000000000000038
[    0.075076] CRIT: [appelfloader] Stack:
[    0.075406] CRIT: [appelfloader]  100017f760  00 00 00 00 00 00 00 00  |........|
[    0.076043] CRIT: [appelfloader]  100017f768  a1 01 00 00 00 00 00 00  |........|
[    0.076680] CRIT: [appelfloader]  100017f770  48 bd 7b 00 10 00 00 00  |H.{.....|
[    0.077324] CRIT: [appelfloader]  100017f778  e0 92 7a 00 10 00 00 00  |..z.....|
[    0.077959] CRIT: [appelfloader]  100017f780  48 bd 7b 00 10 00 00 00  |H.{.....|
[    0.078596] CRIT: [appelfloader]  100017f788  90 f8 17 00 10 00 00 00  |........|
[    0.079233] CRIT: [appelfloader]  100017f790  38 00 00 00 00 00 00 00  |8.......|
[    0.079870] CRIT: [appelfloader]  100017f798  72 fc 41 00 10 00 00 00  |r.A.....|
[    0.080504] CRIT: [appelfloader] Call Trace:
[    0.080871] CRIT: [appelfloader]  [0x000000100041fdcb]
[    0.081311] CRIT: [appelfloader]  Bad frame pointer

It looks like the instance exited fatally.  To see more details about why, run:

        kraft cloud instance get http-python312-hb7ij

Use the recommended command for detailed output. This yields output like:

(ansi)
 
             uuid: 988c3054-09c0-48ef-ac02-877f8352d93f
             name: http-python312-hb7ij
             fqdn:
     private fqdn: http-python312-hb7ij.internal
       private ip: 172.16.3.1
            state: stopped
          created: 2024-04-18T18:24:12Z
          started: 2024-04-18T18:24:12Z
          stopped: 2024-04-18T18:24:13Z
      start count: 1
    restart count: 0
 restart attempts:
     next restart:
   restart policy: never
      stop origin: initiated by kernel (----k)
      stop reason: Out of memory. Try increasing instance's memory (see -M flag). (i127 PGFAULT ENOMEM)
    app exit code:
            image: http-python312@sha256:8b92ed612f8450de94355a0a5b8917a710fd6c902c4a7be80ffb3989b5a99023
           memory: 128 MiB
             args: /usr/bin/python3 /src/server.py
              env:
          volumes:
          service:
        boot time: 96.03 ms
          up time: 166ms

This output shows the stop reason. Here the cause is insufficient memory.

If the stop reason lacks detail, enable debug tracing for the instance.

To do that, in your app of choice in the examples repository, or in an app directory you created, update the runtime entry in the Kraftfile to reference the debug build of the image you use by adding -dbg to the name of the runtime. For example, if you want to run the http-go1.21 example with debug output, update its Kraftfile as follows:

Kraftfile(yaml)
 
spec: v0.6

runtime: base:latest-dbg

rootfs: ./Dockerfile

cmd: ["/server"]

That is, change base:latest to base:latest-dbg. Now Unikraft Cloud is ready to re-deploy:

(bash)
 
kraft cloud deploy --name http-go-strace -p 443:8080 .

You can now inspect the logs as before and view system call tracing:

(bash)
 
kraft cloud inst logs http-go-strace
(ansi)
 
[    0.000000] Info: [libkvmplat] Unikraft Telesto (0.16.2~5b96d531)
[    0.000000] Info: [libkvmplat] Architecture: x86_64
[    0.000000] Info: [libkvmplat] Boot loader : unknown-lxboot
[    0.000000] Info: [libkvmplat] Command line: unikraft netdev.ip="172.16.3.4/24:172.16.3.254:172.16.3.254::http-go-strace:internal" vfs.fstab="initrd0:/:extract::ramfs=2:" virtio_mmio.device=4K@0xd0001000:5 -- /server
[...]
epoll_ctl(0x4, 0x1, ...) = 0x0
epoll_ctl(0x4, 0x1, ...) = 0x0
getsockname(fd:3, <out>sockaddr:{sa_family=AF_0xffffffff901f0002, ...}, <ref:0xc000045a9c>10385019274329587824) = OK
accept4(0x3, 0xc000045ad0, ...) = Resource temporarily unavailable (-11)
nanosleep(0x100052f120, 0x0, ...) = 0x0
clock_gettime(CLOCK_MONOTONIC, <out>timespec:{tv_sec=0, tv_nsec=187025020}) = OK
clock_gettime(CLOCK_MONOTONIC, <out>timespec:{tv_sec=0, tv_nsec=187640540}) = OK
set_robust_list(0x10005839a0, 0x18, ...) = Function not implemented (-38)
rt_sigprocmask(0x2, 0x1000583fb0, ...) = 0x0
mmap(va:0x1050000000, 67108864, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS|0x4000, fd:-1, 0) = va:0x1050000000
mprotect(va:0x1050000000, 135168, PROT_READ|PROT_WRITE) = OK
sigaltstack(0x0, 0x1000583158, ...) = 0x0
rt_sigprocmask(0x2, 0x1000583168, ...) = 0x0
gettid() = pid:6
epoll_pwait(0x4, 0x10001803c8, ...) = 0x0
clock_gettime(CLOCK_MONOTONIC, <out>timespec:{tv_sec=0, tv_nsec=191570733}) = OK
epoll_pwait(0x4, 0x100052eb28, ...) = 0x0
nanosleep(0x100052f120, 0x0, ...) = 0x0
clock_gettime(CLOCK_MONOTONIC, <out>timespec:{tv_sec=0, tv_nsec=192861008}) = OK
clock_gettime(CLOCK_MONOTONIC, <out>timespec:{tv_sec=0, tv_nsec=193478343}) = OK

This mechanism works for all apps and runtimes. Contact Unikraft Cloud on the Discord server and include this output if you need more detail.

While you debug an issue you can mitigate crashes by setting a restart policy. For example, use kraft cloud deploy --restart on-failure to have the platform restart the app if it crashes. Find more info on restart policies here.

Learn more

Last modified on
ImagesKubernetes