Comments (6)
Since IPython ignores the crash message, how should we tell it we've shut down?
For real shutdown there is a shutdown request and reply. In case of real full kernel crash, the heartbeat will fail and the frontend is responsible to ask for ask restart (automatic in case of notebook, prompt user for qtconsole)
from ijulia.jl.
@Carreau, I understand that the heartbeat stop will cause a restart. However, it would be nice have a mechanism to report the reason for the crash to the front-end if possible (obviously not possible for a segfault or similar catastrophic error, but we can certainly do it in cases like uncaught exceptions).
from ijulia.jl.
Shouldn't one just catch all exceptions and report them as errors? Are there situations where we can send a message back but can't continue operating? That seems like a strange intermediate situation.
from ijulia.jl.
Sure, that's one possibility, just send a pyerr
message back to the frontend and continue operating. But we may want some mechanism to exit if we get into an infinite loop of exceptions this way.
(For now I want to crash on an unhandled exception just to make sure that we don't accidentally ignore it.)
from ijulia.jl.
The above patch changes the behavior so that it tries to continue on an unexpected exception. However, sending a pyerr
message doesn't seem to work, perhaps because I'm just sending a new UUID as the session
parameter (since this is not in the middle of an execute_request
). @minrk or @Carreau, is there a way to send a pyerr
(or similar) message outside of the context of an execute_request
, to report to the front ends that something may be going wrong in the kernel?
from ijulia.jl.
IURC, the "session" UUID should change only once every kernel restart, it should stay the same across messages.
The only problem is that in between execute request if you publish a message with a message id that no frontend know of, the frontend will just ignore it.
I guess you could try to keep track of last seen session-id and message id by peeking into the messages.
from ijulia.jl.
Related Issues (20)
- Not seeing any output after running plotly() with julia 1.9 HOT 20
- Add precompile statements to reduce startup latency HOT 2
- When `IJULIA_NODEFAULTKERNEL` is defined, `exe` is not. HOT 1
- IJulia is not starting julia 1.9 kernel HOT 4
- UndefVarError: execute_msg not defined HOT 1
- JupyterLab stopped working for Julia 1.9 (404 GETs) HOT 13
- [Feature Request] Capture also input in numbered prompt.
- Debug by cell option for IJulia HOT 1
- Cant connect to kernel after running jupterlab(detach=true)
- error with Julia 1.10 when loading packages that trigger precompilation HOT 5
- Notebook kernel startup/restart is slow
- Kernel exception: usage_request HOT 8
- copy() seems to behave as deepcopy() in IJulia notebook HOT 2
- No plot output from Plots with plotly() backend (incompatible with recent Jupyter Notebook Versions 7.0.0-7.06) HOT 6
- Support parsing of comma separated packages in `] add` HOT 1
- Latex maths not rendered in docstring if function signature is indented by four spaces HOT 2
- cannot build IJulia without disabling Julia's startup.jl HOT 5
- cannot display `body` field of `UnionAll` type in IJulia
- Autocomplete problem in output cell. HOT 4
- notebook() with no X server and elink HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from ijulia.jl.