A VPS can crash even when free -h shows abundant RAM and swapon --show lists unused swap. Those commands report broad guest totals, not every limit that can stop your game server or another workload. A cgroup, systemd service limit, or systemd-oomd policy may restrict the affected process; the failure may instead be disk or inode exhaustion, storage latency, CPU throttling, a network limit, filesystem damage, a kernel panic, or a provider-side reset.
Need a server to try this on? Antihost offers Linux VPS plans with NVMe storage, full root access and a 14-day free trial.
The reliable fix is to identify the failure domain before resizing. Classify what “crashed” means, read evidence from the previous boot, inspect local and provider limits, then change only the limit or component that the evidence implicates.
As an Amazon Associate I earn from qualifying purchases.
Recommended Free Tools
Why comfortable RAM totals do not prove the workload is safe
Host-wide memory is different from an allocation limit
free -h describes memory visible to the VPS as a whole. Your game server may run inside a cgroup, container, or systemd unit with its own ceiling. In cgroups v2, memory.max can be lower than the VM’s apparent capacity, while memory.current shows usage in that group and memory.events records events such as limit hits. A service configured with MemoryMax= has the same kind of local boundary. A process can therefore be killed while unrelated processes still leave plenty of host memory available.
Memory policies can also constrain where an allocation comes from. Linux kernel documentation notes that a memory policy or cgroup can make an allocation fail even when other nodes, or the system overall, have free memory. Always report the domain involved—provider host, VPS, cgroup, service, or container—instead of treating one global number as the diagnosis.
#1 Best Overall
- BLUE LIGHT BLOCKING GLASSES: Our blue light glasses for women boast UV400 protection, blocking harmful blue rays from computer screens. With these bluelight glasses women, you can enjoy gaming, reading, or watching TV with no worries of eye strain.
- EYE RELIEF: Our blue light glasses are engineered to reduce blurry vision or strain caused by prolonged screen time. These womens blue light glasses aid in better sleep by preventing blue light from disrupting your natural sleep cycle, helping you fall asleep faster and enjoy a more restful night's sleep, making them the perfect christmas present.
- COMFORTABLE: Crafted from durable material, our trendy blue light glasses for woman or blue light glasses men ensure long-lasting comfort. The high-transmittance lenses accessories, gamer gifts for men, help keep your eyes relaxed. Perfect stocking stuffers, gifts for gamers men, christmas gifts for men, halloween costumes.
- LIGHTWEIGHT: These clear blue light glasses feature a lightweight frame, meaning they’ll stay comfy for long-term wear. Perfect as blue blockers glasses for women or mens blue light glasses, these non-prescription fake glasses for men can be worn throughout the day. The blue light blocking glasses women feature polycarbonate HD lenses, which are anti-reflective and restore true color.
- WORRY-FREE PURCHASE: If you experience any issues, our support team is ready to assist you quickly. These are the perfect costume glasses for cosplay dressing up.
Swap is a pressure valve, not an immunity shield
Swap provides backing storage for pages that are not resident in RAM; it does not remove cgroup ceilings, repair a full filesystem, or make slow storage responsive. A cgroup can have a separate memory.swap.max, so the guest may show swap while the affected group cannot use it. Heavy swapping can also make a server appear frozen because requests wait on storage rather than completing.
The kernel’s OOM behavior is explicit: “If this is set to 0, the kernel will kill some rogue process, called oom_killer.” That decision can follow a failed allocation in a constrained domain, not a simple test of whether the VPS has any free bytes. Swap exhaustion is one possible contributor, but a local cap, a sudden allocation, or reclaim failure can trigger the same result.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Userspace OOM decisions can precede a kernel OOM
On systems using cgroups v2, Ubuntu’s systemd-oomd uses pressure-stall information and cgroup policy to act before a kernel-space OOM occurs. Its log may name a unit or slice that was terminated while free -h still looked reassuring. Check whether the daemon is installed and active with systemctl status systemd-oomd, then inspect the affected unit’s pressure and memory policy.
First classify what “crash” means
“The VPS went down” describes a symptom, not a mechanism. Distinguish these cases before interpreting logs:
Rank #2
- WINDEX GAMING WIPES: effortlessly clean all your gaming gear and gaming electronics in one easy step with our new Windex Gaming Wipes
- AMMONIA FREE: our ammonia-free formula is safe to use on all of your favorite gaming equipment like consoles, headsets, controllers, screens, keyboards, and more
- PRE-MOISTENED WIPES: each wipe instantly removes dust, dirt, and stubborn smudges and fingerprints leaving a streak-free shine
- ANTI-STATIC: great for cleaning and wiping down your gaming gear and electronics leaving behind no visible residue
- COMPACT RESEALABLE PACKAGE: store your pack of Windex Gaming Wipes anywhere in your home for whenever a quick and easy wipedown is needed
| Observed event | What it usually means | First evidence to collect |
|---|---|---|
| One game server or service exits | The process was killed, failed, or hit its own resource limit. | Unit journal, exit status, OOM and cgroup events. |
| SSH stops responding but the VM remains powered on | A guest freeze, severe I/O wait, network failure, or kernel deadlock is possible. | Provider console, serial output, uptime, kernel log. |
| The guest boots again | A panic, watchdog action, manual reboot, or provider reset may have occurred. | Previous-boot journal and provider event history. |
| No guest evidence survives | The reset may have happened outside normal guest logging, or logs were volatile. | Provider event timeline and serial console. |
Check the provider’s event history beside the guest’s uptime. A host fault or provider reset requires a different remedy from a service that exceeded a cgroup limit.
Evidence-first investigation after a reboot
1. Read the previous boot before changing anything
Run these commands as an administrator:
journalctl --list-boots
journalctl -k -b -1
journalctl -b -1 -p warning..alertThe first lists boots known to the journal; the others show kernel messages and higher-severity entries from the prior boot. If you know the unit, add -u your-service-name. Journal filtering is more useful than scrolling through the current boot because the triggering message often precedes the reboot by seconds or minutes.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →2. Search for signatures, not just the word “crash”
Search the previous-boot output for:
oom-kill,Out of memory,panic_on_oom, orsystemd-oomd.hung_task,BUG:,Oops:, watchdog messages, orKernel panic - not syncing.I/O error,EXT4-fs error,XFS, and filesystem recovery failures.No space left on device, journal write failures, and inode-related errors.
A panic message narrows the branch. Boot guidance from Google Cloud distinguishes memory-exhaustion panic, a crashed init process, filesystem damage, machine checks, and a corrupt kernel image by the text printed with the panic. Capture the exact line, not a paraphrase.
3. Measure every relevant memory boundary
Collect a snapshot while the VPS is healthy and again during an incident when possible:
Rank #3
- Ergonomic & Adjustable Design: Fight "text neck" and improve your posture. Designed for comfort, it offers an adjustable height (7.1in-8.5in) (1.4in adjustable range) and angle (5°-85°) Please note: cell phone stand is only 9 inches tall—the length of a standard pencil
- Without blocking subtitles Perfect for office desk Tiktok Face Time Zoom video—this is the ultimate kitchen desk essential for phones! Perfect viewing experience while reducing neck back strain, A must-have for gamers streamers desk, and anyone who values both style and support!
- Sleek & Stable Design fits Otterbox Case: Say goodbye to bent cables, tangles, and tip-overs! LISEN’s stylish adjustable desk phone holder features a built-in charging hole and smart cable management, keeping your phone secure and your workspace clutter-free cool gadgets for men
- ANTI-SLIP DESIGN free hands: The under 10.00 items easy gadgets back and the bottom of this kitchen essentials for women are fully covered by anti-skid silicone, which LISEN stand up desk decor can provide maximum protection device from any scratches and slides
- Your Perfect Viewing Companion:If your product arrives damaged or breaks within 365 days, reach out to LISEN for a quick and hassle-free solution. Struggling with assembly, like tightening the screw? We’ve got your back! Your experience matters to us. (Product including: phone Holder stand*1, screw*1, hexagon screwdriver*1, specification*1)
free -h
swapon --show
vmstat 1 5
ulimit -a
systemctl status systemd-oomdFor the affected cgroup, inspect memory.max, memory.current, memory.events, and memory.swap.max. Record the cgroup path for the service rather than reading a parent group and assuming it applies to the process. In a container, inspect the container’s limits from the host or its runtime interface as well as the values visible inside the container.
4. Check storage and inode capacity
Run:
df -h
df -iA filesystem can have free gigabytes but no free inodes, preventing new files, sockets, PID files, or journal entries. Conversely, a full volume can stop databases, save files, package operations, and even the journal itself. Review kernel messages for I/O errors and compare disk latency and IOPS with the provider’s metrics. Do not assume that adding RAM can restore write capability.
5. Correlate CPU, load, and I/O measurements
Use several short samples rather than one percentage:
iostat -xz 1
pidstat -d 1
mpstat -P ALL 1Microsoft’s troubleshooting guidance emphasizes relationships between metrics, not isolated utilization figures. High load with relatively low CPU use points toward I/O wait or blocked tasks. CPU that is not saturated while latency rises suggests storage or network trouble. CPU that repeatedly rises and falls under a steady workload can indicate a burst allowance or provider throttling. Compare these samples with provider-side CPU-credit, IOPS, throughput, and network graphs when those are exposed.
Rank #4
- Easy Installation: Our gaming lights have been designed to enhance your entertainment experience. Simply position the led light bar next to your TV or game console to create a game room environment with ease. They come with USB power and a 60 inch cord for direct mounting and mobility
- Audio Syncing & Extensive Mode: The built-in high-sensitivity music sensor allows you to sync the TV led lights to your favorite music. Lights react and move to the beat of music, film and game audio, creating an incredible immersive sound and light experience – and all without the need for an additional hub. You can choose from 64+ patterns, including strobe, illusion, jump, energy, gradient, running, chasing, fluttering mode etc. Each mode can be applied to different scenes
- 64+ Chasing IC+RGB & W Effect: The independent control design of IC+RGB and white all-in-one multifunctional rgbw led bar lights monitor backlight allows you to easily switch the led lights for desk between IC+RGB and white to enjoy a variety of lighting effects, you can choose white lighting mode with high brightness for better illumination when working, etc. Choose IC+RGB lighting mode when you need to create an entertaining atmosphere for the light show, 16 million colours of Color changing rgb light bar let you dance to the beat and create a dynamic party atmosphere
- Dynamic and Dimmable Lights: 1%-full brightness free brightness adjustment gives you the flexibility to adjust dimmable led light bar for room decorations or a variety of occasions
- USB Powered: USB light bars for room can be powered by powernank, usb charger, TV, Monitor, laptop PC mainframe, making behind TV led lighting bar easy to install in any location.TV led lights are powered by the TV usb interface / PC mainframe, can be turned on/off synchronously with TV / PC
6. Preserve logs across the next failure
If journald is only stored in volatile memory, a reboot can erase the evidence you need. Create the persistent journal directory and configure bounded retention in /etc/systemd/journald.conf. The controls SystemMaxUse, RuntimeMaxUse, SystemKeepFree, and RuntimeKeepFree let you cap journal growth while reserving space for the rest of the system. After changing the configuration, restart systemd-journald and verify that entries are being written under /var/log/journal. Choose limits that fit the volume; an unbounded journal can create the disk-full condition you are trying to diagnose.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →7. Compare guest evidence with provider evidence
If the guest is unreachable and there is no previous-boot journal, open the provider’s serial console or emergency console. Check event history for host maintenance, forced resets, storage incidents, and network actions. A provider reset or host fault will not necessarily leave a normal Linux log entry. Save console output and timestamps so they can be correlated with your monitoring data.
Failure signatures and the next check
| Signature | Likely domain | Next check |
|---|---|---|
oom-kill or Killed process |
Kernel OOM or a constrained memory domain. | Victim process, cgroup path, memory.events, memory.max, and swap limits. |
systemd-oomd entries |
Userspace pressure-based termination. | PSI data, unit policy, and the cgroup selected by the daemon. |
Kernel panic - not syncing |
OOM panic, hung task, machine check, NMI, crashed init, filesystem, or kernel-image failure. | Exact panic reason, prior-boot journal, and serial-console output. |
No space left on device or full inode count |
Capacity or inode exhaustion. | Largest consumers, deleted-but-open files, inode-heavy directories, and safe filesystem expansion. |
| High load with low CPU and rising wait | Storage latency, queueing, or IOPS throttling. | iostat -xz 1, provider disk metrics, queue depth, and write pattern. |
| Guest unreachable with no guest logs | Provider reset, host fault, or panic outside SSH visibility. | Provider event history and serial console. |
How non-memory bottlenecks make a VPS look crashed
Disk latency and I/O throttling
A game server that saves world state, writes logs, or loads assets can block on storage even when CPU and RAM are plentiful. Sustained queue depth, high await times, or a provider IOPS cap can make SSH time out and watchdogs fire. Check whether latency rises at the same time as saves, backups, log rotation, or database checkpoints. Reducing write bursts, moving oversized logs, or selecting storage with an appropriate performance limit addresses this domain; adding swap does not.
Filesystem corruption or a full boot volume
An unclean shutdown, full disk, or underlying I/O error can damage the filesystem and prevent a normal boot. Google Cloud describes a damaged boot filesystem as commonly following those conditions. Treat repeated filesystem errors as a repair and data-integrity incident: preserve an image or snapshot when possible, use the provider’s recovery environment, and follow the filesystem’s repair procedure rather than repeatedly rebooting a damaged volume.
Best Value
- Compatible with Wide Screens - To ensure compatibility with the dual monitor mount, your each monitor must meet three conditions at the same time: First, computer screens size range: 13 to 32 inches. Second, screen weight range: 4.4 to 19.8 lbs. Third, the back of the monitor screen must have VESA mounting holes with a pitch of 75x75mm or 100x100mm.
- Regarding the compatibility with desks - Your desk must meet three conditions at the same time: First, desk material: Only wooden desks are recommended, plastic or glass desks cannot be used. Second, desk thickness range: 0.59" - 3.54". Third, the bottom of the desk should not have any cross beams or panels, as this will interfere with installation. We recommend carefully checking that your desk and monitors meets all above conditions before purchasing.
- Dual C-Clamp Hold - Worried your dual monitors might wobble or slip? Our upgraded base uses a larger platform plus a dual C-clamp structure to lock the dual monitor arm firmly to your desk. Each arm safely keeps your screens steady while you type, click and game—no shaking, no sliding, just a clean and secure setup you can trust every day. It also provides Grommet Mounting installation choice, both options ensure stable and secure fixation for your 0.59" - 3.54" desk.
- Full-Motion Adjustment For Comfortable View - Pull the screen closer when you’re deep in a spreadsheet, push it back to watch videos, or rotate to portrait for coding — moving everything smoothly with just one hand. The monitor stand offers +85°/-50° tilt, ±90° swivel and 360° rotation. Raise your monitor up to 15.75″ to support a healthy sitting posture. Whether you’re working from home, gaming through the night, or switching between video calls and documents, getting the screens to your natural line of sight helps relieve neck, shoulder and back strain so you can stay focused longer with less fatigue.
- Keep Your Desk Organized: By lifting both screens off the desktop, this dual monitor stand opens up valuable space for your keyboard, notebook, docking station or a simple, clutter-free work area. Built-in cable management guides wires along the arms, keeping cords out of sight and out of the way. Enjoy a tidy, modern workstation that looks as good as it feels to use.
CPU throttling and burst limits
A VPS can become unresponsive when a provider enforces a CPU quota or a burst allowance expires. Look for periodic drops in delivered CPU alongside rising request latency, not just a high average CPU percentage. Compare guest mpstat samples with the provider’s throttling or credit metrics. If the limit is confirmed, reduce concurrency, move the workload to a plan with sustained CPU, or ask the provider for the relevant quota; do not infer a RAM shortage from the symptom.
Network ceilings and external dependencies
A server may be healthy but appear down when its network throughput, packet rate, firewall, or upstream dependency fails. Rising latency with unsaturated CPU and normal disk metrics warrants network graphs, interface errors, connection counts, and provider events. Keep an out-of-band console path so an SSH or game-protocol outage does not remove your only diagnostic channel.
Choosing a remedy without making the incident worse
| Evidence | Targeted remedy | Risk and reversibility |
|---|---|---|
| Guest-wide memory pressure, no tighter local cap | Reduce workload peak or increase RAM; review swap placement and performance. | Resize may require downtime; workload reduction is usually reversible. |
memory.max, service MemoryMax, or swap cap reached |
Raise or remove the local limit deliberately, or split the workload. | Reversible configuration change, but higher limits can move pressure elsewhere. |
systemd-oomd termination |
Correct the unit’s pressure behavior and workload peak after confirming policy. | Disabling protection blindly can turn a contained kill into a guest-wide OOM. |
| Full disk, inodes, or journal | Free verified-safe data, rotate logs, expand the filesystem, and bound retention. | Deletion can cause data loss; snapshot or copy important data first. |
| I/O errors or filesystem damage | Use recovery tools, provider diagnostics, and a verified backup or replacement volume. | Repair and migration can require downtime; repeated writes may worsen damage. |
| CPU, disk, or network throttling | Change workload shape or provider capacity after confirming the metric. | Plan changes may cost more and may require migration. |
| Provider reset or host fault | Supply timestamps and console evidence to the provider; prepare failover. | Guest configuration cannot repair an external host event. |
A RAM upgrade is rational only when the evidence shows a guest-wide memory shortage. The decision should account for the failure domain, quality and persistence of evidence, recovery time, data-loss risk, reversibility, and whether the provider exposes the CPU, I/O, burst, and event metrics needed to verify the result.
Common investigative mistakes
- Assuming “free RAM” means every service can allocate it. Check cgroup and unit limits first.
- Adding swap without checking
memory.swap.max, disk latency, and filesystem capacity. - Reading only the current boot. A reboot can hide the triggering kernel messages.
- Treating high load as CPU saturation without checking I/O wait and queue depth.
- Running
echo 3 > /proc/sys/vm/drop_cachesas a cure. Kernel documentation warns that dropping caches can add CPU and I/O cost and is intended for testing or debugging, not routine recovery. - Deleting logs or game data during an incident before taking a snapshot or recording what was removed.
A practical incident checklist
- Record the outage time, last known-good time, uptime, and provider events.
- Decide whether a process died, the guest froze, it rebooted, or the provider reset it.
- Save
journalctl --list-boots,journalctl -k -b -1, and the previous boot’s warning-and-alert output. - Search for OOM, panic, watchdog, filesystem, I/O, and no-space signatures.
- Capture
free -h,swapon --show,vmstat 1 5,df -h,df -i, andulimit -a. - Inspect the affected cgroup’s memory files and the service’s systemd limits.
- Correlate
iostat,pidstat, andmpstatwith provider CPU, disk, network, and event metrics. - Preserve persistent, bounded journald storage before the next reboot.
- Apply one domain-specific change, then monitor whether the corresponding signature disappears.
When increasing VPS size is the right answer
Resize only after the measurements show sustained guest-wide memory pressure or a provider capacity limit that your workload genuinely needs. If the victim is always one service inside a capped cgroup, changing the VPS size may leave the cap unchanged. If the evidence points to disk latency, inodes, a panic, or a provider reset, more RAM changes none of those conditions. The fastest durable fix is the smallest change that removes the demonstrated constraint while preserving logs and a recovery path.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
