JuanJo Ciarlante jjo

UM Cloud — Optimization Pass (late May 2026)

Two-day stabilization + scale-up campaign covering Ceph health, the OpenStack control plane (rook-ceph, RabbitMQ, MariaDB, memcached, Horizon, nova/neutron/cinder), and a reproducible load-test harness. End state: cluster healthy, 40 concurrent VM creates verified at 2m10s–2m17s with 0 errors, no MessagingTimeouts in the control plane.

1. Ceph — recovery + resilience

Acute fires

rabbitmq-rabbitmq-0 stuck mount (CSI op-lock leftover from um-kros-03 hard reboot). Fix: restart csi-rbdplugin DaemonSet pod on the node + csi-rbdplugin-provisioner rollout.

Summary:

how	walltime	CPU	MEM	obs
build-c-native/main	0:01.10	99%	1408k
build-c-wasm/main.wasm	0:07.26	99%	37888k	with `wasmer`
build-c-wasm/main.wasm	0:07.34	99%	15232k	with `wasmtime`
build-go-native/main	0:05.80	100%	1536k
build-go-wasm/main.wasm	0:05.61	99%	55388k	with `wasmer`
build-go-wasm/main.wasm	0:05.69	97%	28768k	with `wasmtime`

Building a sinusoidal, back from now, using vector and label_replace.

round(
    8  # <- number of nodes
    *
    (1-sin(vector((time()-(${__to:date:seconds}))/(3600*10) * pi()))) / 2
)
* on() group_right() (
 label_replace(vector(0.0445000) , "spend", "cpu", "", "") * 8 # &lt;- vCPU per node

	#!/bin/bash
	# ZFS Performance Tuning Script
	# Run with appropriate privileges (sudo/root)

	set -e # Exit on any error

	# Check if ZFS module is loaded
	if ! lsmod \| grep -q zfs; then
	echo "Error: ZFS module not loaded"
	exit 1

	# Transform a map of
	# { <wave1>: [<cluster1>, ... <clusterN>], ... }
	# to entries as
	# absent(__foo{cluster="cluster1", wave="wave1"}) or
	# absent(__foo{cluster="cluster2", wave="wave1"})
	# which are join-able via promQL against cluster- labeled timeseries
	$ ./scripts/monitoring/rightsizing/clusters-by-wave.sh karpenter \| jq -r '. as $r
	\| ["ops","prod0","prod1","prod2","prod3","prod4"] as $wave
	\| reduce $wave[] as $w ([]; . + (( $r[$w] // [] )
	\| map({wave:$w, cluster:.})))

	//@version=5
	indicator("jjo's TBO inChart (similar to CBT's)", shorttitle="jjo-TBO-inChart", overlay=false)

	// === INDICATOR INPUTS ===
	// Core TBO Settings
	tbo_fast = input.int(9, "TBO Fast Period", minval=5, maxval=20, group="TBO Core")
	tbo_medium = input.int(21, "TBO Medium Period", minval=15, maxval=35, group="TBO Core")
	tbo_slow = input.int(50, "TBO Slow Period", minval=30, maxval=100, group="TBO Core")
	tbo_smoothing = input.int(3, "TBO Smoothing", minval=1, maxval=10, group="TBO Core")

	---
	apiVersion: apiextensions.k8s.io/v1
	kind: CustomResourceDefinition
	metadata:
	annotations:
	controller-gen.kubebuilder.io/version: v0.16.5
	name: netfilterclasses.netfilter-exporter.grafana.com
	spec:
	group: netfilter-exporter.grafana.com
	names:

	#!/usr/bin/env python
	## Ejemplo minimo para resolver caso1: solo dos sucursales,
	## se debe generalizar teniendo por ej suc_count=6 via codigo.
	## --jjo, 2025-04-25

	from mininet.net import Mininet
	from mininet.node import Controller, RemoteController, OVSController
	from mininet.node import CPULimitedHost, Host, Node
	from mininet.node import OVSKernelSwitch, UserSwitch
	from mininet.node import IVSSwitch

	# Use Debian Bookworm as the base image
	FROM debian:bookworm-slim

	# Update the package list and install necessary dependencies
	RUN apt-get update && apt-get install -y \
	firefox-esr \
	libcanberra-gtk3-module \
	libdbus-glib-1-2 \
	libgtk-3-0 \
	libx11-xcb1 \

	### Add below stanza to your ~/.ssh/config
	## Parsed as vmname.REGION.PROJECT.gcp
	Host ..*.gcp
	ProxyCommand gcloud beta compute ssh --tunnel-through-iap $(echo %h\|awk -v FS=. '{ printf ("--project=%%s --zone=%%s %%s", $3, $2, $1) }') --command="nc -q0 localhost 22"