weaveworks / scope Goto Github PK

Monitoring, visualisation & management for Docker & Kubernetes

Home Page: https://www.weave.works/oss/scope/

License: Apache License 2.0

Shell 6.66% Go 60.86% Makefile 0.81% HTML 0.03% JavaScript 28.33% Python 1.57% C 0.12% HCL 0.96% Dockerfile 0.55% Starlark 0.11%

visualization monitoring microservices containers docker kubernetes ecs mesos

scope's Issues

JS error after nodes delta received

Somehow I seemed to have arrived at an inconsistent topology: theinternet was in the removal list, but another node that was not to be removed still had it in its adjacency list.

Safari was still connected to it:

Usually, Safari would have been in the list of modified nodes. But that didnt happen here.
Is there any chance we could blame the websocket connection? Then, I need to make the topology code in JS more robust and add some sanity check if all nodes are still there.

Stop embedded the static UI content in a massive go file.

We should just stick this in the container and serve it from there.

Hide pseudo nodes

I'm not sure phantom is the right term, but these are connections for which we don't have information about one end:

Right now, we should all the connections. Its only three above, but I've seen it be as many as 10, and its tends to dominate the view then.

We could:

Hide connections unless we see both ends
Group them together into on (we have a "The Internet" blob - why are these not grouped into that?)
Group them by IP
Leave the current behaviour

I'm leaning towards (1), but could be persuaded by (2) or (3). What do you guys think?

CircleCI + Coveralls

—

Automate the process for building the UI

And have it built in the standard weave make all commands (and by travis, circle etc)

Embed changeset id and log it on startup

Take new screenshot that includes window chrome for readme

I think it might look nicer with some sort of border; perhaps the browser window will be enough?

Also, centre align it pls.

Seems to be missing a repo.

Trying to build and can't get past this:

probe/main.go:17:2: cannot find package "github.com/weaveworks/procspy" in any of:
    /usr/local/go/src/github.com/weaveworks/procspy (from $GOROOT)
    /src/src/github.com/weaveworks/procspy (from $GOPATH)
make: *** [build] Error 1

There's no repo there as far as I can tell. Am I doing something wrong?

Method for deduping reports

With #13, the possibility of an app connecting to a probe multiple times has been introduced. This will allow an app to receive the same report more than once, and potentially double count. We need to disallow this.

Options:

Make report merge fully idempotent ie r1.merge(r1) == r1 and so on
Make probes only accept one connection per app
Include a random unique id in each report, and dedupe on the app

I like (1), as it has nice properties for the future; but it is also the hardest.

Consider using docker events API instead of polling in docker process mapper.

Rebase away the history

—

Hostname in scope container is "scope"

Get coverage metrics for client code

Clicking "Applications" twices cause the screen to go blank

Scope doesn't see example app when running in a container.

This is pretty serious...

Minimize JS bundle

Using webpack maybe?

Scope should use fewer docker layers

vagrant@ubuntu-14:~/src/github.com/weaveworks/scope$ ./scope launch
Unable to find image 'weaveworks/scope:latest' locally
Pulling repository weaveworks/scope
8987330a95fa: Pulling image (latest) from weaveworks/scope
8987330a95fa: Pulling image (latest) from weaveworks/scope, endpoint: https://registry-1.docker.io/v1/
8987330a95fa: Pulling dependent layers
511136ea3c5a: Download complete
a8be93256def: Download complete
2db66c93c72e: Pulling metadata
2db66c93c72e: Pulling fs layer
2db66c93c72e: Download complete
98b0c7b63452: Pulling metadata
98b0c7b63452: Pulling fs layer
98b0c7b63452: Download complete
22618500139b: Pulling metadata
22618500139b: Pulling fs layer
22618500139b: Download complete
bd7d0b16def3: Pulling metadata
bd7d0b16def3: Pulling fs layer
bd7d0b16def3: Download complete
6ff49fd79084: Pulling metadata
6ff49fd79084: Pulling fs layer
6ff49fd79084: Download complete
4371bc9dee57: Pulling metadata
4371bc9dee57: Pulling fs layer
4371bc9dee57: Download complete
9819431aea96: Pulling metadata
9819431aea96: Pulling fs layer
9819431aea96: Download complete
02289b89fa78: Pulling metadata
02289b89fa78: Pulling fs layer
02289b89fa78: Download complete
281e7cdd2aae: Pulling metadata
281e7cdd2aae: Pulling fs layer
281e7cdd2aae: Download complete
8987330a95fa: Pulling metadata
8987330a95fa: Pulling fs layer
8987330a95fa: Download complete
8987330a95fa: Download complete
Status: Downloaded newer image for weaveworks/scope:latest
bc267a103940127e67b2dda613f77a80400c104c5ecb08c2c5bf185215ff50bf

Update README for container-related stuff

Make scope work _over_ the weave network

It doesn't mix with --net=host

Docker process mapper

In the probe, extract metadata from the Docker runtime, and attach it to nodes as appropriate.

Usable marketing website

Fill with proper copy
Update fonts
Generate screenshots from new UI
Screencasts?

Source https://github.com/weaveworks/website/blob/master/scope, target http://weave.works/scope.

Details pane doesn't grow or scroll when there's a lot of information in it

If there are multiple tables in the details pane, it exhibits overflow: visible; behavior. I guess we want overflow: scroll;.

Get per-link bandwidth info

An option would be to make the weave router track this for (src, dst mac addr pairs) and provide an api to support it.

Investigate using errcheck

https://github.com/kisielk/errcheck

Reduce topology selectors to 2

Applications (current processname)
Hosts (current networkhost)

I'm trying to think about how to get the word Container in there, still...

Make scope stop quicker

vagrant@ubuntu-14:~/src/github.com/weaveworks/scope$ time ./scope stop

real     0m10.197s
user     0m0.061s
sys      0m0.076s

Don't use bash as your entrypoint

Its doesn't handle signals properly:

https://blog.phusion.nl/2015/01/20/docker-and-the-pid-1-zombie-reaping-problem/

Add "container" topology

Group together processes in the same container.

Stop using supervisord in the container

Its appears to be the reason the container is so big. There must be a more lightweight version.

Wrapping of container id in details panel

Re-introduce the group toggle button

Would be a nice value-add. No huge priority.

Processes with the same cmd line are grouped together

For instance, is I symlink app and probe from something called 'run', and run them both, I see a connection from run to run...

/home/weave # ps auxf
PID   USER     TIME   COMMAND
    1 root       0:00 /sbin/runsvdir /etc/service
    7 root       0:00 runsv probe
    8 root       0:00 runsv app
    9 root       0:00 ./run
   11 root       0:05 ./run
   28 root       0:00 /bin/sh
   36 root       0:00 ps auxf
/home/weave # ls -al /proc/9/
total 0
dr-xr-xr-x    9 root     root             0 May 19 13:22 .
dr-xr-xr-x  118 root     root             0 May 19 13:22 ..
dr-xr-xr-x    2 root     root             0 May 19 13:25 attr
-rw-r--r--    1 root     root             0 May 19 13:25 autogroup
-r--------    1 root     root             0 May 19 13:25 auxv
-r--r--r--    1 root     root             0 May 19 13:25 cgroup
--w-------    1 root     root             0 May 19 13:25 clear_refs
-r--r--r--    1 root     root             0 May 19 13:25 cmdline
-rw-r--r--    1 root     root             0 May 19 13:22 comm
-rw-r--r--    1 root     root             0 May 19 13:25 coredump_filter
-r--r--r--    1 root     root             0 May 19 13:25 cpuset
lrwxrwxrwx    1 root     root             0 May 19 13:25 cwd -> /etc/service/app
-r--------    1 root     root             0 May 19 13:25 environ
lrwxrwxrwx    1 root     root             0 May 19 13:25 exe -> /etc/service/app/app
dr-x------    2 root     root             0 May 19 13:22 fd
dr-x------    2 root     root             0 May 19 13:25 fdinfo
-rw-r--r--    1 root     root             0 May 19 13:25 gid_map
-r--------    1 root     root             0 May 19 13:25 io
-r--r--r--    1 root     root             0 May 19 13:25 latency
-r--r--r--    1 root     root             0 May 19 13:25 limits
-rw-r--r--    1 root     root             0 May 19 13:25 loginuid
dr-x------    2 root     root             0 May 19 13:25 map_files
-r--r--r--    1 root     root             0 May 19 13:25 maps
-rw-------    1 root     root             0 May 19 13:25 mem
-r--r--r--    1 root     root             0 May 19 13:25 mountinfo
-r--r--r--    1 root     root             0 May 19 13:25 mounts
-r--------    1 root     root             0 May 19 13:25 mountstats
dr-xr-xr-x    6 root     root             0 May 19 13:25 net
dr-x--x--x    2 root     root             0 May 19 13:25 ns
-r--r--r--    1 root     root             0 May 19 13:25 numa_maps
-rw-r--r--    1 root     root             0 May 19 13:25 oom_adj
-r--r--r--    1 root     root             0 May 19 13:25 oom_score
-rw-r--r--    1 root     root             0 May 19 13:25 oom_score_adj
-r--r--r--    1 root     root             0 May 19 13:25 pagemap
-r--r--r--    1 root     root             0 May 19 13:25 personality
-rw-r--r--    1 root     root             0 May 19 13:25 projid_map
lrwxrwxrwx    1 root     root             0 May 19 13:25 root -> /
-rw-r--r--    1 root     root             0 May 19 13:25 sched
-r--r--r--    1 root     root             0 May 19 13:25 schedstat
-r--r--r--    1 root     root             0 May 19 13:25 sessionid
-r--r--r--    1 root     root             0 May 19 13:25 smaps
-r--r--r--    1 root     root             0 May 19 13:25 stack
-r--r--r--    1 root     root             0 May 19 13:25 stat
-r--r--r--    1 root     root             0 May 19 13:25 statm
-r--r--r--    1 root     root             0 May 19 13:25 status
-r--r--r--    1 root     root             0 May 19 13:25 syscall
dr-xr-xr-x    9 root     root             0 May 19 13:25 task
-r--r--r--    1 root     root             0 May 19 13:25 timers
-rw-r--r--    1 root     root             0 May 19 13:25 uid_map
-r--r--r--    1 root     root             0 May 19 13:25 wchan

Proper page titles

For better browsability in history.

Error building /app

Not sure if you wan't bug reports or not yet... but when running make build I had an issue. Drilling down a bit the error occurs when building in /app

scope/app# go build -v
github.com/gorilla/context
github.com/gorilla/mux
github.com/gorilla/websocket
github.com/weaveworks/scope/report
github.com/weaveworks/scope/xfer
_/root/scope/app
go build _/root/scope/app: /usr/local/go/pkg/tool/linux_amd64/6g: signal: killed

I'm running on ubuntu 14.04 using go version go1.4.2 linux/amd64

Integration tests

./scope launch the container, do a http get
./scope launch on two containers with weave installed, check they can talk to each other

we could use the bash "micro" framework weave uses, and all the gce scripts to bring up vms.

Make fails cryptically if docker-squash is not installed

Removing intermediate container 0786e37577cb
Successfully built bfaf16cdf9ad
docker save weaveworks/scope:latest |   -t weaveworks/scope | tee scope.tar |  docker load
/bin/sh: 1: -t: not found
FATA[0000] Error response from daemon: EOF

Makefile should check that all dependencies exist, and produce actionable error messages if something needs to be installed.

top - 17:10:03 up 6 days, 21:50,  3 users,  load average: 0.23, 0.58, 0.74
Tasks: 111 total,   2 running, 109 sleeping,   0 stopped,   0 zombie
%Cpu(s):  9.9 us, 17.9 sy,  0.0 ni, 71.8 id,  0.0 wa,  0.3 hi,  0.0 si,  0.0 st
KiB Mem:   1017796 total,   895976 used,   121820 free,    17096 buffers
KiB Swap:  3063804 total,    50512 used,  3013292 free.   124512 cached Mem

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND                                                                                                       
27243 root      20   0   14936   8480   2424 S  44.3  0.8  18:49.89 probe                                                                                                         
    9 root      20   0       0      0      0 S   1.3  0.0  15:53.28 rcuos/1                                                                                                       
  673 root      20   0 1438672  26436   6108 S   1.3  2.6  36:50.98 docker                                                                                                        
    7 root      20   0       0      0      0 R   1.0  0.0  16:58.19 rcu_sched

ugh ~/src/github.com/weaveworks/scope (enhanced-details) make static
mkdir -p client/dist
docker run -ti -v /Users/peter/src/github.com/weaveworks/scope/client/app:/home/weave/app \
                -v /Users/peter/src/github.com/weaveworks/scope/client/dist:/home/weave/dist \
                weave/scope-ui-build
Unable to find image 'weave/scope-ui-build:latest' locally
Pulling repository weave/scope-ui-build
FATA[0004] Error: image weave/scope-ui-build:latest not found
make: *** [client/dist/scripts/bundle.js] Error 1

./run: exec: line 3: /home/weave/probe: not found
./run: exec: line 3: /home/weave/app: not found
./run: exec: line 3: /home/weave/probe: not found
./run: exec: line 3: /home/weave/app: not found
. . .

Clicking topology that is already displayed results in empty screen

view topology Applications
click on Applications

BOOM! Emtpy screen.

weaveworks / scope Goto Github PK

scope's Issues

Recommend Projects

Recommend Topics

Recommend Org