Steven E. Harris comments

Results 239 comments of


                                            Steven E. Harris

K/V v2: add suport for lease_duration using ttl key

@briankassouf, I don't think this is a duplicate of #6184. At present, if one runs _vault kv patch ttl=60s_, you wind up with a value `vault.Secret.Data["data"]["ttl"] = "60s"`, which is...

otelhttp: Bazel fails to build due to io_opentelemetry_go_otel_internal_metric

I spent many hours today writing patches for Gazelle to apply to fix this problem and several others like it. The patches work, to a point, but the _real_ solution...

Provide ability to Load an Existing Secret or Create a New Secret to Datadog.

If we could mount a Kubernetes _Secret_ into the collector pod as a volume, we could then define environment variables using "spec.containers.env.valueFrom.secretKeyRef" and "spec.containers.envFrom."

Allow parsing YAML documents (and streams) like we do with the "from-json" function

> Was there a problem you recently dealt with that would have benefited from a shell (Elvish or otherwise) being able to read YAML and convert it to the shell's...

Allow parsing YAML documents (and streams) like we do with the "from-json" function

As [I wrote above](https://github.com/elves/elvish/issues/1155#issuecomment-695368428): > I could use an external tool to convert YAML to JSON and then use Elvish's from-json function, but I'm left to wonder why we have...

Prometheus plugin causes much longer serving duration for concurrent HTTP requests

In one Kubernetes cluster suffering this problem, we have 428 services and 1,633 routes. I understand that yielding occasionally during the scrape might help, but the real problem here is...

Prometheus plugin causes much longer serving duration for concurrent HTTP requests

I have been experimenting with a modified copy of the Prometheus plugin, testing my hypotheses about the cause of this problem. It turns out that I was wrong on several...

Prometheus plugin causes much longer serving duration for concurrent HTTP requests

One incremental approach I neglected to mention is to shard or partition the single shared memory zone used for the Prometheus metrics dictionary. Choose a fixed split factor such as...

Prometheus plugin causes much longer serving duration for concurrent HTTP requests

Thank you, Enrique. We have been working with Kong account representatives and engineers since I filed this issue, so we are in good hands there. I filed it publicly deliberately...

Broken cache paths

I noticed the same problem today running _glide update_ with Glide version 0.12.3. ``` [ERROR] Could not find /Users/seh/.glide/cache/src/https-golang.org-x-sys/unix: open /Users/seh/.glide/cache/src/https--Users-seh-.glide-cache-src-https-golang.org-x-sys-unix: no such file or directory [ERROR] Dependency golang.org/x/sys/unix (/Users/seh/.glide/cache/src/https-golang.org-x-sys/unix)...