little bugs with "restart"

44 views
Skip to first unread message

Pierre Guillot

unread,
Nov 16, 2016, 9:28:15 PM11/16/16
to sage-cloud
Easy bug to reproduce under SMC:

--open new jupyter notebook
--choose sagemath 7.3 kernel
--enter 2+2 in first cell, evaluate
--choose "clear all output and restart"

The output is NOT cleared... also there's often a message about the kernel being dead.

Generally speaking, but this is harder to reproduce, the "restart" functionality often crashes (or freezes) everything. Usually clicking on "reconnect" fixes the problem, but that's weird, and sometimes it doesn't work either.

Just reporting this...

Pierre

William Stein

unread,
Nov 16, 2016, 11:18:05 PM11/16/16
to sage-cloud
On Wed, Nov 16, 2016 at 6:28 PM, Pierre Guillot
<pierre....@gmail.com> wrote:
> Easy bug to reproduce under SMC:
>
> --open new jupyter notebook
> --choose sagemath 7.3 kernel
> --enter 2+2 in first cell, evaluate
> --choose "clear all output and restart"
>
> The output is NOT cleared... also there's often a message about the kernel
> being dead.
>

I was only able to reproduce this once (the first time), and all other
times is worked fine. However, if I disabled my network the above
error would always happen. I think it's the expected behavior for
Jupyter if the network isn't very fast. You should be able to
replicate it every time by disabling your networking before clicking
"clear all output and restart". Jupyter devs often assume Jupyter
is running locally on a laptop and that the network is super fast.

> Generally speaking, but this is harder to reproduce, the "restart"
> functionality often crashes (or freezes) everything. Usually clicking on
> "reconnect" fixes the problem, but that's weird, and sometimes it doesn't
> work either.

I've not been able to replicate this. Does the notebook content
matter? Which web browser and OS are you using?

-- William

>
> Just reporting this...
>
> Pierre
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-cloud" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-cloud+...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-cloud/4bd4657f-6dab-4d5f-a5e8-b7e350fa701c%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.



--

Best Regards,
William Stein

CEO, SageMath, Inc.

Peter Luschny

unread,
Nov 22, 2016, 4:49:18 AM11/22/16
to sage-cloud
PG> Generally speaking, but this is harder to reproduce, 
PG> the "restart" functionality often crashes (or freezes) 
PG> everything. Usually clicking on "reconnect" fixes the problem, 

I can confirm this observation. I have reported it in similar form in
https://groups.google.com/d/msg/sage-cloud-members/vm-b6PJz_DE/JAiuu8sBAQAJ
and also in help requests directly on SMC.

WS> Jupyter devs often assume Jupyter is running locally on a 
WS> laptop and that the network is super fast. 

I do not know what 'super fast' is. 
My network guarantees a 100Mbit/s connection. 

Peter

William Stein

unread,
Nov 22, 2016, 1:02:45 PM11/22/16
to sage-cloud
At least 1000Mbits/s with very low latency.

> My network guarantees a 100Mbit/s connection.
>
> Peter
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-cloud" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-cloud+...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-cloud/2c67dfe3-2e2b-46f1-9a69-e4d9c9e494dd%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages