XIO

xWire

Posts from every feed QUEST has found across the web — blogs, news, podcasts, channels and more.

Recent Questions - Server Fault

most recent 30 from serverfault.com

Architecture Review: Concurrently testing 10 Wi-Fi CPEs using 10x USB Wi-Fi 6 dongles on a single Linux host

We are designing an automated test bench to test returned customer premises equipment (CPE / Wi-Fi routers—specifically Huawei LG8245X6-50 and ZTE ZXHN H3600P). The bench currently handles a pool of 45 physical modem slots with 10 parallel test workers. Our current goal is to validate both Wi-Fi bands (2.4 GHz and 5 GHz) on up to 10 modems simultaneously. We want to review our proposed hardware/software design and ask for feedback or alternative architectural suggestions. Current Proposed…

hbing.uk company let allow IP spoof, illegal?

I bought a dedicated server from this provider: https://hbing.uk/dedicated-servers — D6. The server is very powerful, with a full-duplex 20 Gbps connection. We use it to sell VPS services to our clients. Unfortunately, one client has repeatedly used our VPS service to launch DDoS attacks using DNS, LDAP, SSDP, and other protocols. I banned the client, but they kept coming back with multiple accounts and purchasing our VPS service again. So far, I have banned the same client around 8 times, and…

Request for Advice on GPv1 to GPv2 Storage Upgrade and Backup Plan

We currently have two Blob Storage accounts that store information related to a VM with an attached database. This VM is critical to our workflow, so we would like to proceed with the storage upgrade from GPv1 to GPv2 carefully. Before performing the upgrade in the production environment, we would like to conduct some tests to confirm that there will be no impact on the VM, database, or related operations. In addition, we would like to establish an alternative or rollback procedure in case any…

Central PKI Azure

Has anyone implemented a centralised PKI where the Root CA is stored in Azure Key Vault and workloads can automatically obtain the Root CA when they are deployed, as well as receive updated versions when the Root CA needs to be rotated or renewed? I’m looking to implement a centralised PKI for an Azure hub-and-spoke architecture, where workloads across the spokes can automatically obtain and trust the Root CA. The workloads could include VMs, AKS, containers, App Services and Azure Functions.…

I am deploying a private Cloud Run background-worker service in europe-west1 for a controlled production rollout

Our public Vision API creates Google Cloud Tasks for internal background work. Each task should call the worker’s /tasks/dispatch endpoint using a dedicated service account and an OIDC token; the worker must remain private and accept invocations only from that task identity. Before enabling any real workload, we are trying to verify the deployment with a safe health request to /healthz. A private Cloud Run worker is Ready and has a correct service-level roles/run.invoker binding for a Cloud…

ZFS backups, what is needed to restore?

I have a home server which I use for my personal needs and my company. I have few databases and a Nextcloud instance running on it. Roughly, the setup is following: 1 SSD which is used for "system" with three partitions: EFI FAT partition /boot on a ZFS pool root on a ZFS pool with encryption 3 HDD on raidz1 pool also with encryption, which are used for Nextcloud data + backups of databases As an example, the data pool had been created with this command: zpool create \ -o ashift=12 \ -o…

Windows 2019's Settings apps collapses on open and the search at the Start menu doesn't work

I have around 25 Windows Server 2019 VMs. The have been all working fine. Starting from July or August I started noticing that the Settings apps collapses on open and the search at the Start menu stops working after installing Windows updates. Uninstalling the updates is not an option for me as all of them now are security patches. I also noticed that some servers already have the Settings app and search broken before I do any recent updates. I am not sure since when they stopped working. I ran…

What is the difference between "Eviction" and "Pod Disruption" in Kubernetes?

The Kubernetes documentation defines "Eviction" as: Eviction is the process of terminating one or more Pods on Nodes. "Pod Disruption" is defined as: Pod disruption is the process by which Pods on Nodes are terminated either voluntarily or involuntarily. What I understand from this is that Evictions occur when a Kubernetes component (a controller, kubelet, etc.) terminates one or more pods on nodes as part of its intended functionality. In other words, Evictions are a feature. Whereas, Pod…

Running Podman rootless on a hardened Alpine image

While attempting to get Podman running on a Docker hardened Alpine image: dhi.io/alpine-base:3.24-alpine3.24-dev The image builds, then produces errors like: Error: preparing container aa77c0fc8a287df8a9fd421a2bf01eaa5244d76a4009b188336d7a762075dd91 for attach: crun: mount `proc` to `proc`: Operation not permitted: OCI permission denied Using runc also fails. Current Dockerfile: FROM dhi.io/alpine-base:3.24-alpine3.24-dev RUN apk --no-cache add buildah fuse-overlayfs iptables podman skopeo \ &&…

Physicaly transfer ZFS pools between systems

I do not currently have a system in this condition. My previous experience is with Enterprise HBA based RAID and I'm trying to understand what my blind spots are when moving to ZFS by using scenarios I successfully solved with the HBA previously. Since the HBA managed the volumes I had a lot of flexibility on swapping built arrays between systems. The Array information was separately stored in both the array members and on the controller. This also made it possible to 'SneakerNet' entire…

While creating kubernetes master node I have the following an issue concerning port 10250 that is is already occupied

I ran the following command to create kubernetes master node ~$ sudo kubeadm init --pod-network-cidr=192.168.0.0/16 While that I had multiple issues, after I succeeded, I have collected the token & complete output of the above command. When check 6443 port - 400 http error came, Please find the following Screenshot. As it is 400 it is somewhat positive. And while trying kubeadm join 192.168.29.95:6443 --token o0koxr.yzht29874x22g17d --discovery-token-ca-cert-hash, it's looks like already token…

MS SQL backup to rustfs gives Operating system error 12175

I have Docker on Windows 11, MS SQL, nginx as proxy and rustfs. I have the below folder structure, docker-compose.yaml, nginx.conf, mssql/Dockerfile, mssql.conf, SQL command run via DBeaver and finally the Python code which triggers the backup. I am facing Operating system error 12175. Can anyone help solve this error? Folder structure C:\>tree docker-setup /A /F Folder PATH listing for volume Windows Volume serial number is 000001DD 0EB2:7F97 C:\DOCKER-SETUP | docker-commands.sh |…

How to install GoDaddy SSL certificate files on Debian Apache2?

I have been using GoDaddy for SSL certificates for years, I had the procedure all written down. Now, they have changed the format of the files they send. Here is the list of files in their zip file: -rw-rw-r-- 1 elson elson 1366 Aug 4 00:01 pico-systems.com-root.pem -rw-rw-r-- 1 elson elson 2247 Aug 4 00:01 pico-systems.com-intermediate.pem -rw-rw-r-- 1 elson elson 1975 Aug 4 00:01 pico-systems.com-cross.pem -rw-rw-r-- 1 elson elson 2581 Aug 4 00:01 pico-systems.com-certificate.pem -rw-rw-r-- 1…

How can I run a container without warning about runAsNonRoot user with non-numeric user?

How can I run a container without warning about runAsNonRoot user with non-numeric user? I found a solution setting securityContext with runAsUser 1000. I want to avoid this on a namespace and run containers in that namespace without a warning or aborting. I tried setting the following labels to a namespace but it didn't work. I tested on minikube on Windows 10, however the minikube addon shows that pod security policy can't be enabled for minikube due to it is deprecated.…

Postfix & Rspamd (dkim_signing module) multiple domains how to select correct domain to sign outgoing messages

I have a Postfix server for multiple domains using virtual_alias_domains and virtual_alias_maps. Receiving email works. I want to use Rspamd to add DomainKeys signatures for outgoing messages. Using Thunderbird as the mail client, I want to configure it with multiple "Identities" for a single "account" - where each identity might be any email address mapped in the Postfix virtual_alias_maps - including email addresses at more than one domain. Concrete example: Assume a user called bob can…

Mail deliverability issues: reputation or p=reject?

I am a small business owner with a relatively new .com domain, and I use Google Workspace for my mail. I have a problem with emails from me going into recipient spam folders, however delivery to Gmail inboxes works. At first, DMARC, DKIM and SPF were not configured, and I sent a few mails during that time. I recently configured them and verified with a different gmail address that in the head, they all got a PASS. Today I learned that someone with an Outlook mail received my mail in their spam…

Dovecot closes client connection after successful login, so IMAP session fails

Trying to bring up a Dovecot 2.4.3 installation on my publicly accessible home server. I had a working 2.3.x installation and have followed all the migration advice on the dovecot website. Eventually, I archived the previous config and started again with a de minimis configuration from the dovecot website (1, below). I narrowed the problem down to the server closing the connection immediately after a successful login, but the question is, why does it do that? I used 'openssl s_client' to test…

Postfix trying to deliver mail using old MX entries

We switched some of our company mailboxes (one domain) to Outlook servers. After that our emails are requiring up to 30 minutes to be delivered from our server to Outlook. Checked logs and there is something strange - before delivering almost all of e-mail to outlook there are few tries which ending with "Connection timed out". When I was checking this IP with telnet on server or locally - this IP's are not active and I'm getting timeouts. Tried to force flush DNS cache every minute on server -…

Map uid/gid of mount to specific uid/gid in kubernetes

Kubernetes supports user namespaces, and apparently supports id-mapped mounts. Tools like podman allow for using id-mapped mounts via the idmap option to --mount, but there doesn't appear to be anything in the kubernetes docs about how to use id-mapped mounts. What are the options that are required to be set in the PV/PVC/Pod config for performing an id-mapped mount?

Best way to write ExecStop for multiple child processes (systemd)?

I've written my first systemd user service and I'm wondering what the best way is to kill it. When the script starts it up the processes look like this (output from systemctl status): ... CGroup: /user.slice/user-1000.slice/user@1000.service/app.slice/myservice.service ├─2203135 /bin/sh /home/.../start-myservice-systemd ├─2203136 /home/.../python3 /home/.../bin/the-server --port 8200 └─2203137 /usr/bin/multilog s1000000 n10 /home/.../logs When stopping the service I want all of the processes to…

Puppet fails with Cannot allocate memory - fork(2) on Debian Trixie VMs (Ganeti) unless RAM is increased to 8 GB

I am facing a memory-related issue on Debian Trixie VMs running on Ganeti. These VMs are used exclusively as PostgreSQL database servers. The same Puppet configuration works fine on Debian Bullseye and Bookworm, but consistently fails on Trixie. Environment Hypervisor: Ganeti Guest OS: Debian Trixie VM RAM: 4 GB (fails), works only at 8 GB Workload: PostgreSQL + Puppet agent Puppet version: Puppet 7 PostgreSQL version: 14 Problem When running Puppet (runpuppet), I get multiple failures like:…

Vault Agent Injector: How to render secrets to a path without overwriting existing files there?

I'm trying to deploy PgAdmin4 to Kubernetes. I have a custom config_local.py that contains an oauth2 client secret. I have the contents of config_local.py stored in Vault. Ideally, I'd have the following annotations in my spec.template.metadata.annotations: vault.hashicorp.com/agent-inject-secret-config_local.py: <Path in Vault> vault.hashicorp.com/secret-volume-path-config_local.py: /pgadmin4 vault.hashicorp.com/agent-inject-template-config_local.py: | {{`{{ with secret "<Path in Vault>" }} {{…

Shared secrets with CSI secret sync enabled

I would be interested to understand how to handle shared secrets used by many resources (e.g. deployments) by using CSI Secret Provider Classes. In this moment I have many Helm Releases in the same namespace that references and sync the same secret from Azure Key Vault. By seeing the yaml of one secret referenced more than once, I see many owner references, obv the ReplicaSets of all helm releases's revisions. kind: Secret metadata: creationTimestamp: "2026-01-20T17:22:24Z" labels:…

vSphere Standard 8.0.3 to vSphere Hypervisor 8.0U3e "upgrade" as an option?

Long story short is over the years we've been forced to move from vSphere Essentials to Standard, and now with Broadcom's ridiculous licensing hikes, to Enterprise. So our upcoming renewal is going to increase our costs by more than fivefold. We have a single dual-CPU host. We don't require all the bells and whistles of HA, clustering, vSAN, etc. I do see Broadcom has a free vSphere Hypervisor available. Surprisingly enough --> Free Version. Wondering if we install this for our environment,…

In MacOS Tahoe and Sequoia, who do you trigger the Local Network Access popup from the command line?

We're starting to see a problem where ssh tunneling stops working in MacOS Sequoia and/or Tahoe. The problem stems from MacOS adding "Local Network Access" permissions, which prevent applications from connecting to anything on the LAN unless granted permission. The way it is supposed to work is that when any application does try to access the local network, the OS adds that application to the list of applications that can talk to the local network, but toggled off, and the dialog asks if you…

Redirect no HTTPS Site A to HTTPS Site B

I have this rule set in Apache: RewriteEngine ON RewriteCond %{HTTP_HOST} http://www.siteA.co.uk$ [NC] RewriteCond %{HTTP_HOST} http://SiteA.co.uk$ [NC] RewriteCond %{HTTP_HOST} ^SiteA.co.uk$ [NC] RewriteRule ^(.*)$ https://www.SiteB.co.uk/$1 [R=301,L] http://siteA site redirects perfect, but the http://www.siteA keeps giving me Forbidden you do not have permission

Rename or Migrate a Microsoft Domain Name/Active Directory from one name to another

We have the following problem (names have been changed to protect the innocent): 400 node network virtually all of it Windows 11 Pro machines Multiple servers including an Exchange 2016 server The company registered "ubuntu.org" about 30 years ago, used ubuntu.org on their internal Microsoft Domain, then someone forgot to pay the renewal and a speculator bought it. The company then registered "ubuntu.com" and uses that on all of their external stuff. They configured their Exchange server to…

Docker nftables forward rules seem ignored in the context of policy drop despite of priority

Context I let docker (29.1.4, build 0e6fee6) produce its own rules { "firewall-backend": "nftables" } and I am working on their cohabitation with my main pretty restrictive table table inet filter { chain input { type filter hook input priority filter; policy drop; iifname "lo" accept ct state established,related accept icmp type echo-request limit rate 5/second burst 5 packets accept tcp dport 22 accept comment "Accept SSH IN" .... log prefix "nftables-INPUT: " counter packets 0 bytes 0 drop }…

How can I configure an Azure OpenAI resource so that it allows requests from certain IPs?

I am setting up an Azure OpenAI resource and I need to restrict access so that only requests from our office static IPs and our application server are allowed. By default, the resource seems to be open to "All networks." How can I configure an Azure OpenAI resource so that it allows requests from certain IPs? E.g., openai.com supports allowlists. From https://platform.openai.com/settings/organization/security/ip-allowlist: