profile
viewpoint
Josh Gavant joshgav Discover Financial Services Chicago, IL

joshgav/azure-log-processor 2

Receive and process logs from Azure resources via Event Hubs.

joshgav/azure-scripts 2

Scripts for deploying to Azure platforms.

joshgav/gdrive 2

CLI for Google Drive.

joshgav/azure-go-labs 1

Labs demonstrating how to use Go on Azure.

joshgav/chrome-remote-interface 1

Chrome Debugging Protocol interface

joshgav/diagnostics 1

Tracing Working Group

joshgav/go-sample 1

Sample app for tests.

joshgav/api 0

The canonical location of the Kubernetes API definition.

joshgav/ApplicationInsights-node.js 0

Microsoft Application Insights SDK for node.js

joshgav/appsvc-scratch 0

scripts and tests for Azure App Service

push eventjoshgav/joshgav.github.io

Josh Gavant

commit sha 2a0f2e21cd80c86200c1525393e54ec03fc73db6

add container streaming post

view details

push time in 12 days

push eventjoshgav/joshgav.github.io

Josh Gavant

commit sha 616d97f4d801f57779396e29f2d4bbed28a6af0d

add container streaming post

view details

push time in 12 days

push eventjoshgav/joshgav.github.io

Josh Gavant

commit sha 6c2b105683d1c6293b8f3684e8ae9a378dd76933

add container streaming post

view details

push time in 12 days

push eventjoshgav/dotfiles

Josh Gavant

commit sha a395fdc9b73616e49071dd135d3dc9894c17e84a

support bash completion

view details

Josh Gavant

commit sha 6d3c7e018735fb834c9f16ca075741f70e75de09

add alias k=kubectl

view details

push time in 18 days

issue commentopen-telemetry/opentelemetry-java

Shim from OpenTracing impl to OpenTelemetry API

Thanks @winmat, I've indeed been working with AppD folks and they now have a plan for OT support soon. In particular I've been watching the work in progress here on an AppD exporter :+1:.

joshgav

comment created time in 2 months

push eventjoshgav/charts

Josh Gavant

commit sha 1e8688efd23b7f5805054841ca2ffd30cfc476b6

add MS Teams config to kubewatch

view details

push time in 2 months

issue commentbitnami-labs/kubewatch

MS Teams handler not working in Helm chart

@jonbesga I proposed a fix in https://github.com/bitnami/charts/pull/3189

jonbesga

comment created time in 2 months

PR opened bitnami/charts

[bitnami/kubewatch] add MS Teams config

Description of the change

Exposes Microsoft Teams configuration for kubewatch in values.yaml file for ease of configuration.

Benefits

Use bespoke MSTeams support in kubewatch rather than generic webhook support. Yields better output in client.

Possible drawbacks

None.

Applicable issues

None.

Additional information

I followed the pattern from https://github.com/bitnami/charts/pull/3024.

Checklist <!-- [Place an '[X]' (no spaces) in all applicable fields. Please remove unrelated fields.] -->

  • [x] Chart version bumped in Chart.yaml according to semver.
  • [x] Variables are documented in the README.md
  • [x] Title of the PR starts with chart name (e.g. [bitnami/chart])
+11 -1

0 comment

4 changed files

pr created time in 2 months

push eventjoshgav/charts

Josh Gavant

commit sha e758520ac7ea944c512a8c1c89a2dbb93ab6068e

bump chart version

view details

push time in 2 months

create barnchjoshgav/charts

branch : kubewatch-msteams

created branch time in 2 months

issue openedbrigadecore/brigade

loading brigade.json fails in non-root container

The prestart.js script in the worker which preloads dependencies includes logic which copies the original brigade.json file to /brigade.json here. This fails in containers which cannot write to / and leads to the following error. Could you consider copying these files to a world-writable location like /tmp instead, or perhaps to something known to be writable like /vcs? Thanks.

NB: I included in the following output some messages about yarn caches too as FYI; that's another (small) issue which comes up in non-root containers.

warning Skipping preferred cache folder "/.cache/yarn" because it is not writable.
warning Selected the next writable cache folder in the list, will be "/tmp/.yarn-cache-1070090000".                
warning Cannot find a suitable global folder. Tried these: "/usr/local, /.yarn"
internal/fs/utils.js:230                                                                                           
    throw err;                                                                                                     
    ^                                                    
                                                         
Error: EACCES: permission denied, open '/brigade.json'                                                             
    at Object.openSync (fs.js:458:3)                                                                               
    at Object.writeFileSync (fs.js:1283:35)                                                                        
    at createConfig (/home/src/prestart.js:34:10)                                                                  
    at addDeps (/home/src/prestart.js:45:3)                                                                        
    at Object.<anonymous> (/home/src/prestart.js:41:3)                                                             
    at Module._compile (internal/modules/cjs/loader.js:1156:30)                                                    
    at Object.Module._extensions..js (internal/modules/cjs/loader.js:1176:10)                                      
    at Module.load (internal/modules/cjs/loader.js:1000:32)
    at Function.Module._load (internal/modules/cjs/loader.js:899:14)
    at Function.executeUserEntryPoint [as runMain] (internal/modules/run_main.js:74:12) {
  errno: -13,                                            
  syscall: 'open',                                       
  code: 'EACCES',                                        
  path: '/brigade.json'                                  
} 

Output of brig version:

Latest version as of creation of this issue via go get -u github.com/brigadecore/brigade/brig/cmd/brig

Output of kubectl version:

Client Version: version.Info{Major:"1", Minor:"18", GitVersion:"v1.18.4", GitCommit:"c96aede7b5205121079932896c4ad89bb93260af", GitTreeState:"clean", BuildDate:"2020-06-18T02:57:29Z", GoVersion:"go1.14.3", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"16+", GitVersion:"v1.16.2", GitCommit:"a3ec9df", GitTreeState:"clean", BuildDate:"2020-05-04T12:54:43Z", GoVersion:"go1.12.12", Compiler:"gc", Platform:"linux/amd64"}

Cloud Provider/Platform (AKS, GKE, Minikube etc.):

OpenShift on EC2

created time in 2 months

issue closednodejs/diagnostics

Trace module WIP - Request for feedback

joshgav/node-trace

Hi all. I've started designing and implementing a trace module/system for Node and wanted to get your early feedback. The goal for this module is to provide a common way for native and JS module authors to publish and/or subscribe to trace events both for debugging and logging; and to enable publishing traces back and forth to listeners on both sides of the native/JS boundary.

This could integrate with core TRACE_EVENT work in addition to ecosystem modules; for example, the TRACE_EVENT system could publish to the trace module to route traces to registered listeners. In fact I already implemented that elsewhere as a proof of concept.

Architecture

For now a whiteboard of the design follows; I'll work on something more formal.

Trace Architecture

Green outlines are JS modules; blue outlines are native artifacts. Black arrows show the flow of a trace published by a JS module; red arrows show the flow of a trace published by a native module or internal component.

The central classes are the native and JS TraceBroker classes (in the middle column), which receive published traces and publish them to a list of listeners they maintain. In both JS and C++ publishers call Trace/trace or TraceNextTick/traceNextTick to publish a trace. Listeners are function pointers registered via TraceBroker::RegisterListener in C++ and functions registered via EventEmitter.on in JS.

The JS TraceBroker class inherits from events.EventEmitter to handle listener management and event publication.

The native and JS brokers are wired together through the private DispatchJSFromNative and _dispatchNativeFromJS functions respectively.

The example (in ./examples) works and demonstrates the concept of a trace flowing from JS to both JS and native, try the following:

git clone https://github.com/joshgav/node-trace && cd node-trace
npm install
node-gyp rebuild
node ./examples/trace_example.js

But much work remains; tasks next on my list include the following:

  • [ ] Add tests.
  • [ ] Fill in placeholders in the code.
  • [ ] Determine how to best transform C++ maps to JS objects and vice versa and implement.
  • [ ] Add basic filtering in the brokers.

ASK: What do you think of the design? What concerns, e.g. perf, should I be thinking about? What similar modules or attempts exist which I could learn from or work with? And of course contributions are welcome :)

closed time in 2 months

joshgav

issue commentnodejs/diagnostics

Trace module WIP - Request for feedback

Can be closed, too dated to be of use. Thanks!

joshgav

comment created time in 2 months

fork joshgav/kashti

Kashti is a dashboard for your Brigade pipelines.

http://kashti.sh

fork in 2 months

fork joshgav/api

The canonical location of the Kubernetes API definition.

fork in 2 months

startedopen-telemetry/opentelemetry-java-instrumentation

started time in 2 months

more