What people keep running into with server administration
Complaints tagged server administration, each raised by more than one person. One-off posts are not included.
15 recurring problems · 98 people · 5 forums we read
Raised more than once
Most people first.
- 22How to stop excessive bot traffic and AI scrapersMassive volumes of automated scrapers overwhelm sites with tens of thousands of public pages, driving hosting costs up significantly. This overwhelming bot traffic ruins the reliability of web data and forces owners to look for specialized blocking services to regain control.
- 10Why do self-hosted emails keep going to spam?Major email providers constantly flag self-hosted messages as spam, leaving owners unable to reach inboxes reliably. This forces people to abandon self-hosting entirely and rely on external services like Fastmail just to get their messages delivered.
- 10Why is it so hard to manage short-lived server certificates?Short-lived SSL certificates force administrators to manually update keys across dozens of servers every few months. This overwhelming administrative overhead makes routine maintenance unsustainable and leaves systems at risk of frequent disruptions.
- 9Why do frequent security updates make maintaining websites so stressful?Routine security patches can turn an ordinary workday stressful when they hit alongside a steady stream of new vulnerabilities. For sysops and site maintainers managing large fleets on end-of-life PHP versions that cannot simply be upgraded, rapid vulnerability disclosures mean relying on third-party or manual backpatches just to keep systems covered.
- 8Contractors asking for admin access but not knowing basic ITUnqualified contractors frequently request full administrative control over infrastructure while failing to understand basic IT tools like firewalls, virtualization, or web servers. This leaves companies vulnerable because these contractors do not know how their requested changes will impact production systems.
- 8Why are computer fleet updates failing to install?Updates are failing after completely installing and restarting, leaving administrators with persistent error codes across multiple servers and machines. Short of fully reimaging the systems, standard recovery methods are failing to resolve the issue.
- 5Server randomly rebooting with no error logs or dumpsMost unexpected server reboots trigger blank Kernel-Power 41 events with zero bugcheck parameters and leave management controller logs completely silent. Because these hard resets generate no minidumps or error entries, troubleshooting hits a dead end until underlying hardware issues like a failing UPS are identified.
- 4Moving a virtual domain controller without a USN rollbackOfficial guides offer conflicting advice on how to move a virtual machine domain controller to new hardware safely. This leaves administrators stuck guessing whether they have to completely demote and rebuild the DC from scratch.
- 4Can I run a local LLM on an old server?Old enterprise servers often have firmware restrictions that stop the motherboard from recognizing generic GPUs, requiring you to buy special branded hardware. If you try to run local LLMs without a GPU, any dense model over 10B parameters will be too slow. While big MoE models can perform faster despite long loading times, these hardware limitations often make buying new equipment the better choice.
- 4Locked out of a remote server after changing settingsA single remote access change—like switching ports, disabling the network interface, or hitting shut down instead of restart—can drop your connection and block you from getting back in to undo it. Without physical access or someone on-site to intervene, you cannot revert the setting yourself.
- 3How to set up local server replication and redundancySmall MSP environments often lack straightforward tools for full local server replication, forcing technicians to piece together complex solutions without clear best practices. This makes it difficult to design reliable redundancy setups or validate configurations left behind by previous engineers before presenting them to clients.
- 3Website crashed after updating PHP versionUpgrading PHP often breaks websites because older themes, plugins, and custom code become incompatible with the newer version. This mismatch throws fatal errors that make both the front end and the admin dashboard completely inaccessible, leaving site owners unable to fix the issue without manual debugging.
- 3Windows Server freezing at the welcome screenWindows Server 2019 virtual machines randomly stop accepting remote desktop connections and hang indefinitely at the welcome screen or frozen desktop. Even though remote PowerShell commands may still connect, console and RDP logins lock up during the authentication or user settings loading phase. This leaves administrators unable to reach a working desktop session.
- 3Free ways to push updates and force rebootsPushing updates for free often fails because machines will not restart automatically to finish the installation. End users never restart their computers on their own, and servers frequently get stuck in a pending reboot state during scheduled maintenance windows. This forces administrators to manually trigger the reboots via RDP or remote tools just to get the updates installed.
- 2Why did DNS break after promoting a domain controller?Promoting a backup domain controller to primary causes all DNS entries to disappear on the PDC, resulting in a complete failure of name resolution. Network traffic mysteriously fails to hit the second DNS server during this transition, leaving administrators confused.
Comes up alongside
Topics that keep appearing on the same problems — not topics with similar names.