Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

bouncing the probe garbles view #2646

Closed
rade opened this issue Jun 25, 2017 · 4 comments
Closed

bouncing the probe garbles view #2646

rade opened this issue Jun 25, 2017 · 4 comments
Assignees
Labels
accuracy Incorrect information is being shown to the user; usually a bug bug Broken end user or developer functionality; not working as the developers intended it component/ui Predominantly a front-end issue; most/all of the work can be completed by a f/e developer
Milestone

Comments

@rade
Copy link
Member

rade commented Jun 25, 2017

  1. launch scope
  2. launch chrome and point at scope
  3. reset the local view state (just to ensure consistency of test conditions)
  4. go to the process view, which should show something like
    screenshot from 2017-06-25 11-55-04
  5. stop scope
  6. start scope; the display will change to something like
    screenshot from 2017-06-25 11-55-53

Notice that a) this is a mixture of old and new nodes, and b) the topology selector says 'Containers' even though what is shown is processes.

@rade rade added bug Broken end user or developer functionality; not working as the developers intended it component/ui Predominantly a front-end issue; most/all of the work can be completed by a f/e developer accuracy Incorrect information is being shown to the user; usually a bug labels Jun 25, 2017
@fbarl
Copy link
Contributor

fbarl commented Jun 26, 2017

Thanks the for clear instructions on how to reproduce this one @rade - I was noticing a similar behaviour before but never figured there was such a simple pattern to it!

The two issues you mentioned are indeed separate although they share the cause:

  1. Topology selector shows wrong info - I thought this one would have been solved by modifying the line 127 in https://github.com/weaveworks/scope/pull/2600/files#diff-5897bc396ca616224f48758015a6ece1, but it might be that there is still some issues with it or that the scope version that you were using locally was taken before that commit.
  2. The nodes are merged wrongly - here we should probably implement a mechanism that clears up the nodes after the probe reconnects and/or have smarter nodes delta merging.

Now we know how to reproduce the issue, I'll try to solve it one of the following days :)

@fbarl fbarl self-assigned this Jun 26, 2017
@rade
Copy link
Member Author

rade commented Jun 26, 2017

it might be that there is still some issues with it or that the scope version that you were using locally was taken before that commit.

Hmm. I could swear my scope was up to date, but trying to reproduce the issue just now, I only see the 2nd problem.

@rade rade modified the milestone: 1.6 Jul 12, 2017
@rndstr rndstr self-assigned this Jul 12, 2017
@rndstr
Copy link
Contributor

rndstr commented Jul 12, 2017

  1. Topology selector shows wrong info – I can reproduce. This happens because upon restart, the first response for api/topology does not have any nodes at all. So it hides all the nav items except Containers and Hosts, then selects Containers since that's the first one available. It keeps the graph view untouched. The second response does contain the nodes and Processes & co shows up but we keep the selection on Containers.

@rndstr
Copy link
Contributor

rndstr commented Jul 12, 2017

[moved 2. to #2708]

rndstr added a commit that referenced this issue Jul 12, 2017
If the scope-app API unexpectedly restarts it has no report
at hand until it receives one from the probe and sends the
node count 0 for all topologies to the frontend. Once the report
arrives, it will send the proper count.

What happened was the frontend did hide Processes for a short
time till the node count recovered. This moved the topology
selection to the always visible Containers (hide_if_empty == false)
while keeping the graph as is.

Once the node count recovers, Processes comes back but the
selection is still at Containers.

This fixes to keep the selected topology visible at all time
even if the API returns a node count of 0. This allows to
keep looking at the graph. The topology will disappear once
the user selected a different topology while a backend response
is received.

Fixes #2646
rndstr added a commit that referenced this issue Jul 12, 2017
If the scope-app API unexpectedly restarts, it has no report
at hand (until it gets one from the probe) and sends node
count 0 to the frontend for all topologies. Once the report
arrives, it will send the proper count.

What happened was the frontend did hide Processes for a short
time till the node count recovered. This moved the topology
selection to the always visible Containers (hide_if_empty == false)
while keeping the graph as is.

Once the node count recovers, Processes comes back but the
selection is still at Containers.

We now keep the selected topology visible at all time even if
the API returns a node count of 0. This recovers nicely when
the correct node counts come in. Once the user selects a different
topology while and a backend response arrives, it disappears.

Fixes #2646
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
accuracy Incorrect information is being shown to the user; usually a bug bug Broken end user or developer functionality; not working as the developers intended it component/ui Predominantly a front-end issue; most/all of the work can be completed by a f/e developer
Projects
None yet
Development

No branches or pull requests

3 participants