serverfault.com

How can I configure a GCP (Google Cloud Platform) account so that one cannot spend more than some given amount of money on it? For example, https://openai.com/ allows users to set a hard limit:

I'm running a Windows 2012 Server and was wondering if there was some software that essentially took a complete snapshot and backup of the entire server. Something where, should both of my HDs simultaneously fail (I'm running RAID 1 across 2 HDs), I could load up new drives and restore everything completely without having to re-install all of the software separately. I'm running a web server off of it and frankly I've forgotten how to do some of that stuff with IIS, and I'm just thinking proactively about having a complete backup somewhere. I've heard of Norton Ghost and making "image" files, but I never worked with it and I'm not sure if that fits what I'm looking for. Any suggestions?

sudo ceph-volume lvm activate seems not to work on any of my storage nodes: When I issue the command together with the OSD ID 0 and the FSID sudo ceph-volume lvm list gave me, I always get an exception like this: mixtile@blade3n2:~$ sudo ceph-volume lvm prepare --data /dev/nvme0n1[sudo] password for mixtile: --> Device /dev/nvme0n1 is already prepared mixtile@blade3n2:~$ sudo ceph-volume lvm activate 0 0381e029-3f06-4fcb-9580-4c95b110eb86 Running command: /usr/bin/ceph-authtool --gen-print-key Running command: /usr/bin/ceph-authtool --gen-print-key Running command: /usr/bin/chown -R ceph:ceph /var/lib/ceph/osd/ceph-0 Running command: /usr/local/bin/ceph-bluestore-tool --cluster=ceph prime-osd-dir --dev /dev/ceph-58be7e97-a687-4afc-a8ca-312b3590ba78/osd-block-0381e029-3f06-4fcb-9580-4c95b110eb86 --path /var/lib/ceph/osd/ceph-0 --no-mon-config stderr: 2025-12-31T17:39:15.436+0100 ffffab994040 -1 bluestore(/dev/ceph-58be7e97-a687-

I'm trying to lock down my nginx reverse proxy to my local network only for specific domains. I have set the following in /etc/nginx/nginx.conf: http { set_real_ip_from 10.0.0.0/24; real_ip_header X-Forwarded-For; etc } And the following in my sites-available .conf file: server { listen 8443 ssl http2; server_name <domain>; allow 10.0.0.0/24; deny all; location / { proxy_set_header Host $host; proxy_pass http://n.n.n.n:nnnn/; client_max_body_size 0; proxy_connect_timeout 3600; proxy_send_timeout 3600; proxy_read_timeout 3600; send_timeout 3600; } } Using this setup however I get a 403 error when connecting to the domain. Both http_realip_module and realip_module are loaded. Am I missing something? Or doing it completely wrong?

I have an issue where my main desktop PC (running Fedora) is unable to resolve hostnames for machines on the LAN. These are normally provided by a DNS server running on Debian (using bind9/named). When this situation happens, a Wireshark capture shows it tries the lookup on localhost then 224.0.0.252 (which seems to be related to multicast DNS) before giving up and never querying the real server. This can be resolved as a workaround by simply disconnecting and reconnecting to the LAN, but this is somewhat tedious and can only resolve it after the fact (i.e. when a a scheduled backup fails). Hostnames are given as part of the domain zerker.home (i.e. zeta.zerker.home) Here's a proper nslookup: $ nslookup zeta Server: 127.0.0.53 Address: 127.0.0.53#53 Non-authoritative answer: Name: zeta.zerker.home Address: 10.0.1.10 And a bad output:

I'm using mariadb 15.1. I'd like to know whether preparing a backup multiple times is safe or can be dangerous. mariadb-backup --prepare [options] After some searching I have to outcomes: AI says it's "safe but redundant", but doesn't provide any reference to those exact words. In a post somebody stated that it could be dangerous. So, can somebody confirm whether it's safe but redundant [how and why] or it's dangerous [how and why]?

I'm experiencing a significant billing discrepancy on Google Cloud Platform across two of my billing accounts for December 2025. The Usage Reports page shows one amount, but the Transactions/Billing page shows a substantially higher amount. Account 1 (tmcc.in) Billing Account ID: 0190F9-779808-54BA19 Usage Reports show: ₹180,904.72 (Dec 1-29, 2025) Transactions page shows: ₹225,542.07 Unexplained difference: ~₹44,637 Account 2 (tmbc.in) Billing Account ID: 01F3D2-F416B8-9DCEB0 Usage Reports show: ₹8,291.96 (Dec 1-29, 2025) Transactions page shows: ₹29,377.54 (before GST) The billed amount is 3.5x higher than reported usage What I've checked: The Usage Reports are filtered for the same date range GST (18%) doesn't account for this large difference Both accounts experienced sudden

I'm using this script to populate an Allow firewall rule with the latest list of EXO IP addresses. I want to allow port 25 traffic only from those addresses. After implementing the rule, connections from unauthorized IP addresses to my system continue. I created the rule according to these instructions. I created a companion Block rule for all port 25 traffic, in hopes that the two would work together to restrict traffic to only those IPs. However, that blocked everything. I'm running my mail server behind a NAT configuration, if that makes a difference. How do I create a Windows Firewall rule (ruleset?) to allow

I've activated "Stored Completions" in my Azure OpenAI Resource Instance: How can I access "Stored Completions" from Azure OpenAI (Microsoft Foundry)?

I'm looking for a reliable, repeatable way to confirm whether a website is actually down, or if the issue is caused by access restrictions rather than a service outage. Typical scenarios I've encountered: The site times out or fails to load from one network The same site is reachable from other regions or networks DNS resolves correctly, but HTTP requests fail Different HTTP status codes depending on source network Before escalating issues to hosting providers or making DNS / firewall changes, what are the best technical checks to perform to confidently determine: a true service outage vs ISP-level blocking firewall / WAF rules geo-based filtering routing issues What combination of signals (HTTP status, DNS, traceroute, multi-region checks) do you consider sufficient evidence of a real outage?