Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallIf Linux still reports 100% usage after you deleted files, stop deleting and reconcile the accounting first. df reports blocks allocated by a filesystem, while du totals files that are still visible in directory trees. A process can keep a deleted file open, inode slots can be exhausted even when byte space remains, and reserved space or storage layers can make the figures look contradictory.
Need a server to try this on? Antihost offers Linux VPS plans with NVMe storage, full root access and a 14-day free trial.
Use the sequence below to identify the exact filesystem, determine whether bytes or inodes are exhausted, find open deleted files, check journal retention, and only then choose a repair that will actually release space.
As an Amazon Associate I earn from qualifying purchases.
Why Linux can say 100% when cleanup appears to work
A full-disk alert is a filesystem accounting result, not a count of filenames you can see. The first question is therefore whether the shortage is in data blocks, inode records, or a layer that your directory scan does not include.
What df measures
df asks the mounted filesystem how many blocks are allocated, how many are available to the account being used, and what percentage that represents. Its result includes space occupied by files that are reachable from directories, files that have been unlinked but remain open, filesystem metadata, and any blocks reserved or otherwise unavailable to ordinary users. POSIX also permits the reported percentage to exceed 100 when available space becomes negative.
#1 Best Overall
- [Ultra-Lightweight Ear Buds Designed for Small Ears] Each earbud weighs only 3.7g and features a compact, ergonomic in-ear design made especially for small ears. Secure, low-profile, and comfortable for workouts, all-day wear or overnight sleeping.The ultra-slim profile sits flush in your ear with no hard edges poking or rubbing, so you can even sleep on your side without discomfort.
- [Immersive Stereo Sound with TOZO OrigX Technology] TOZO OrigX tuning delivers clear vocals, balanced mids, and natural stereo sound for music, podcasts, and videos.
- [Long Battery Life for Daily Use] Get up to 7 hours of playtime on a single charge, with up to 32 hours total using the charging case—ideal for workdays, commuting, and extended listening sessions.
- [Bluetooth 5.3 & Stable Connection] Bluetooth 5.3 provides fast pairing, stable wireless performance, and reduced dropouts as you move around home or office.
- [Deep Bass with Clear Vocals] High-performance drivers produce punchy bass while keeping vocals clean and detailed for everyday listening.
What du measures
du walks directory entries and adds the blocks attributed to files it can reach. An unlinked file has no directory entry, so it is invisible to a normal du walk even though its blocks remain allocated. A recursive walk can also cross into other mounts unless you restrict it, causing its total to be unrelated to the filesystem reported by df.
How to read the mismatch
| Observation | Most useful interpretation | Next check |
|---|---|---|
df and du are close |
Visible files account for most allocated blocks. | Find the largest directories and decide what can be archived, moved, or removed. |
df is much higher than du |
Look for open deleted files, another mount boundary, snapshots, overlay storage, or filesystem metadata. | Run sudo lsof +L1 and inspect the mount layout. |
df -i is 100% while byte space remains |
Inode records, not data blocks, are exhausted. | Locate directories containing huge numbers of small files. |
| Usage is high only for one path | The path may be on a separate filesystem with its own limit. | Run df -hT /that/path and use that mount point for all later checks. |
Run a safe diagnostic sequence
Capture the outputs before changing anything. Keep a copy of df -hT, df -i, a one-level directory summary, and the open-file listing. Those records make it possible to verify that a change fixed the intended filesystem rather than merely moving the symptom.
Recommended Free Tools
1. Identify the filesystem and type
Start with the path that triggered the alert:
df -hT /path
The -h option makes sizes readable, and -T adds the filesystem type. Note the device, mount point, total, used, available, and percentage columns. Repeat the command for any path involved, such as /var, a database directory, or a container data directory. If the paths resolve to different mount points, diagnose each filesystem separately.
2. Check inode capacity separately
Run:
df -i /path
Compare the inode percentage with the block percentage. A filesystem can have gigabytes free yet reject new files when its inode table is full. Conversely, deleting a few large files will release blocks but may do almost nothing for an inode shortage caused by millions of tiny files.
3. Compare blocks with visible directories
Use a filesystem-bounded summary at the mount point shown by df:
sudo du -xhd1 /mountpoint
The -x option prevents the walk from entering other filesystems. Sort the displayed directories mentally or with your usual shell tools, then repeat the command on the largest directory. Continue one level at a time rather than launching an unrestricted scan of the entire root tree on a distressed server. A large result in /var may be logs, caches, package data, databases, or container layers; ownership must be established before removal.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →4. Look for deleted files still held open
When the df total is substantially larger than the filesystem-bounded du total, run:
Rank #2
- 【IP68 Waterproof Multifunctinal Smart Watch】Torjalph fitness activity tracker smart watch is IP68 waterproof technology ,you can wear it when swimming,handwashing,it's rain-proof which means that you perform multiple exercise modes outside without worrying about sweat or rain corroding your fitness smart watch,and this watch equipped with a 230 mAh large capacity battery magnetic charging ,and this sports activity tracker watch can be used 5-7 days If fully charged.
- 【Larger 1.83'' Full Screen Watch & Multi-dial Selection】1.83 inch full HD color touch screen smart watch for men / women provides excellent 240*284 resolution image quality ,This Bluetooth smart watch offers a better touch experience and remarkable immersion and ,more than 200 watch faces for you to choose from App"Da Fit ", you can change the dial style according to your mood to wear keeping the world on your wrist
- 【Accurate Activity Tracker Smart Watch 】sports watch activity tracker uses sweat and dust resistant breathable silicone strap, our Fitness Tracker Smart Watch for Android iOS Phones supports 12 sports modes,walking,running,bicycle,swimming,badminton,basketball,football,rope skipping. This bluetooth calls smart watch for men women is lightweight and stylish, bringing more convenient and relaxing sports experience while also providing more protection.
- 【Your 24 Hour Personal Daily Health Monitor】Fitness watch for women men full touch screen IP67 waterproof bluetooth smartwatch for Android iOS Phones with heart rate blood pressure monitor multifunction Andriod smart watch with steps counter sleep monitor sports fitness watch, monitoring blood oxygen saturation allows you to understand your physical condition in real time to track fitness progress ,to know your health status in time.
- 【More Functions Bluetooth Call Smart Watch】This fitness activity smart watch full screen for iPhone and Android phones includes a bunch of new features including:The bluetooth smart watch can hang up the phone, receive and display WhatsAPP /Text /Facebook /Twitter /Messenger reminder.Support sedentary reminder, alarm clock, stopwatch, timer,weather information,music control, remote camera control compatible with Android 5.0 and above or iOS 9.0 and above smartphones.
sudo lsof +L1
On systems where that expression is unavailable, use:
sudo lsof | grep deleted
Record the command name, process ID, user, file descriptor, and size. A nonempty result is a lead, not permission to terminate the process. Several entries can refer to the same process or file, and some entries are harmless short-lived descriptors.
5. Check which service owns each process
For every substantial entry, map the PID to its service before acting. Depending on the distribution, useful checks include:
ps -fp PIDto see the complete command line.systemctl status service-nameto confirm the unit and its state.readlink -f /proc/PID/exeto identify the executable.
Do not infer ownership from a filename alone. A log may be opened by a supervisor, a worker, or a helper process, and restarting the wrong component can interrupt unrelated work.
6. Measure systemd journal usage
On systems using systemd-journald, run:
journalctl --disk-usage
Compare the journal total with the du results and inspect the retention settings in journald.conf. The relevant controls are SystemMaxUse=, RuntimeMaxUse=, SystemKeepFree=, and RuntimeKeepFree=. The current manual documents defaults of 10% and 15% for the two keep-free controls, with each calculated default capped at 4 GiB. These controls govern retention; they are not a substitute for identifying an already-full filesystem.
7. Verify mount and storage boundaries
Use the filesystem type from df -T and inspect the layout with your distribution’s mount tools. A generic du / can cross into separate mounts, while a container’s overlay filesystem may present a merged view that does not correspond to the host’s apparent directory size. Snapshots and copy-on-write metadata can consume space without appearing as ordinary files under the mounted directory.
8. Decide whether the shortage is genuine
Only after the preceding checks should you classify the incident as visible data, open deleted files, inode exhaustion, journal retention, reserved space, or a filesystem-specific layer. That classification determines whether a restart, targeted cleanup, snapshot operation, or capacity change is appropriate.
Rank #3
- [Low Noise High Fidelity] New bee guitar cord with 22AWG oxygen-free-copper center conductors provides enhanced signal transfer, the PVC insulation material and the high-density stranded copper double shield provide noise rejection for quiet operation
- [Ideal for Choice] New bee 1/4 straight and right angle metal connectors, ensures a secure connection in any instrument, pedal, or amp jack. L shape great fro on stage guitar stands
- [Wide Compatible] Suitable for mixer, power amplifier, microphone, effector, decoder, equalizer, electoronic organ, electric guitar, electric bass and other electronic equipments. Widely used in the installation of smart home, karaoke speaker, professional stage audio, etc
- [Easy Disassemble] Making incidental repairs quick and easy! Very nice, flexible cable
- [Dont Hesite to Order] 24 hours customer services and professional technology team are standing by
Release space held by an open deleted file
Removing a pathname unlinks the directory entry; it does not invalidate file descriptors already held by a running process. The inode and its data blocks remain allocated until the last descriptor closes. Red Hat describes this behavior for Linux and Unix systems, and Alibaba Cloud documentation notes that deleted files still consuming space are absent from ordinary directory inspection. This is the classic reason for a high df total paired with a much smaller du total.
Read the lsof result carefully
An entry commonly shows a PID, process name, user, file descriptor, file type, device, size, and a path ending in (deleted). The displayed size is a useful estimate, but several descriptors may point to one file and sparse files may not consume their apparent length in blocks. Concentrate on the filesystem identified earlier and on entries large enough to explain the difference.
Close the descriptor through the owning service
The safest release is a normal close performed by the application. Use the service’s documented log-rotation or reload operation when it supports one. If a reload does not close the descriptor, schedule a controlled restart:
- Confirm the unit name and maintenance window with
systemctl statusand the service owner. - Use the service’s normal stop or restart procedure rather than sending an arbitrary signal.
- Wait for the process to exit and for its replacement to become healthy.
- Run
df -hT /mountpointagain and verify that the expected blocks were released.
A reboot also closes every descriptor, but it is a broad action that interrupts all services and can hide the actual owner. Treat it as a last-resort maintenance choice, not as the first cleanup command. Never kill a PID merely because its deleted-file entry is large; confirm its role and the operational impact first.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
When the space does not drop after a restart
Re-run sudo lsof +L1. A supervisor may have started a replacement while another worker still holds the original file, or a different process may be responsible. If no open deleted file remains, return to mount boundaries, snapshots, and filesystem metadata rather than repeating restarts.
Manage journald without damaging active logs
Journal files are managed data, not disposable text files. Deleting active files by hand can confuse journald, remove evidence needed for incident analysis, and fail to release the blocks you expect. Use the journal’s own reporting and distribution-approved vacuum commands.
Understand the retention controls
SystemMaxUse= and RuntimeMaxUse= cap persistent and runtime journal storage. SystemKeepFree= and RuntimeKeepFree= reserve free space for other workloads. The documented defaults for the keep-free settings are 10% and 15% of the containing filesystem, with each default calculation limited to 4 GiB. Explicit size values and other configuration entries can change the effective behavior, so inspect the active configuration rather than assuming the defaults apply.
Rank #4
- Digital Family Calendar: BSIMB digital calendar wall touch screen puts an end to your chaotic life. Designed for busy families, the digital calendar keeps you organised and in step with your family. Assign different colours and tasks to each family member to make family scheduling easier and more intuitive.
- Easy Setup and Auto-Sync: Plug in, connect to WLAN and auto-syn your electric calendar in minutes. Digital wall calendar compatible with Google calendar, iCloud calendar, Outlook calendar, Cozi calendar and Yahoo calendar. Add events directly on the device or through the ecalendar app. Electronic calendar stays current whether you're at work or traveling—no manual updates needed.
- All-In-One Home Assistant: Smart calendar builds habits and encourage independence with personalised chores. Display daily meal plan to reduce the daily ‘What's for dinner?’ stress. To-do lists keep track of everything you want. Turning chores into valuable achievements with stars and rewards, making life easier and more enjoyable.
- Manage Anytime Anywhere: BSIMB electronic calendar for family stay connected with the powerful ecalendar app. Access and manage your digital wall calendar anytime, anywhere with free features like stars & rewards, photo screensaver, chore charts, meals plans, to-do lists etc. Electronic planner paid features like Magic Import tailor advanced tools for users with special needs.
- Stay in Touch with Loved Ones: Share photos or videos from your phone to electronic wall calendar with app, instantly update your life moment, and keep in touch with your family, especially those who are far away; you can also surprise your smart calendar digital wall by sending him a wish card; it's also great for keeping tech-illiterate elders in touch with the youngsters.
Vacuum only after measuring
Use journalctl --disk-usage to establish how much space is involved, then apply a retention rule approved for the machine. Common policies vacuum by age or total size, but the exact command and policy should match your distribution and compliance requirements. Afterward, check df again. The journald manual warns that when a filesystem later fills, journald may stop adding data without immediately removing every existing archive; a full disk therefore may require an explicit, policy-driven vacuum.
Free tools Windows power users keep installed
One-click scans. No signup required.
Prevent a repeat incident
Set limits that leave room for databases, package operations, and emergency diagnostics. Recheck the settings after upgrades and confirm that persistent storage is on the filesystem you intended. Retention controls prevent future growth; they do not repair inode exhaustion or release blocks held by an unrelated deleted file.
Fix inode exhaustion
Inodes store file metadata and directory references. A workload that creates enormous numbers of small files can consume every inode while leaving substantial byte capacity unused. Symptoms include failures to create temporary files, package-install errors, or service errors that mention no space even though df -h still shows available gigabytes.
Confirm the scope
Use df -i on each affected mount. Then perform a filesystem-bounded inventory of directories, for example:
sudo find /mountpoint -xdev -type f | wc -l
Use more targeted counts under the largest directories, and include directories themselves when investigating applications that create deep trees. Counts alone do not identify safe deletions; inspect ownership, age, naming patterns, and the application’s retention policy.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Remove or consolidate safely
- Use the application’s cleanup or rotation feature for caches, queues, and temporary artifacts.
- Archive batches of small files into an appropriate store only when the application can read that format.
- Delete files through the service’s documented retention procedure, then verify that no process still has them open.
- Do not delete database internals, mail queues, container layers, or lock files solely because they are numerous.
Deleting a few large files addresses blocks, not inode records. To resolve an inode shortage, the number of directory entries must fall or the filesystem must be recreated or expanded with an inode layout suitable for the workload.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Investigate mounts, containers, snapshots, and filesystem metadata
Separate mounts
A directory such as /var/lib may be a mount point of its own. Space consumed there will not be freed by deleting files from the parent filesystem, and a root-level du can either include it or skip it depending on options. Always use the mount point printed by df -T and keep du -x in the comparison.
Container overlay storage
Overlay filesystems combine a read-only lower layer with a writable upper layer. A deleted file in a container, a stopped container’s writable layer, or an image cache can consume host storage outside the path you first inspected. Identify the container runtime’s data mount and use its supported prune or retention operation. Do not remove files directly from an overlay upper directory while the runtime is using it.
Snapshots and copy-on-write filesystems
Snapshots preserve old extents after the active file is changed or deleted. Btrfs subvolumes, snapshot tools, and quota groups require filesystem-specific inspection; XFS and other filesystems have different metadata and quota behavior. A snapshot can keep blocks allocated even when the live directory tree is small. List snapshots with the filesystem’s supported tooling, confirm which ones are required for rollback or backup, and remove only an approved snapshot.
Reserved blocks and metadata
Some filesystems reserve blocks for privileged recovery and metadata operations. Those blocks can make ordinary-user availability lower than the raw free-block count. Do not change reservation settings merely to make a percentage look better. First confirm the filesystem type, the purpose of the reservation, and the operational margin required for repairs and logging.
Choose a remedy by what it changes
| Remedy | Releases already allocated blocks? | Service restart needed? | Data-loss risk | Addresses bytes or inodes | Scope |
|---|---|---|---|---|---|
| Close or restart the service holding an open deleted file | Yes, when the final descriptor closes | Often, unless a supported reload closes it | Low when graceful; workload interruption is possible | Bytes | Any filesystem; diagnosis depends on lsof |
| Journal vacuum under an approved policy | Yes, for archives removed by the policy | Usually no | Loss of historical logs according to the selected retention | Bytes, not inodes | systemd-journald |
| Remove application caches or temporary files | Yes, after descriptors close | Sometimes | Ranges from negligible to severe if the path was misidentified | Bytes and, for many files, inodes | Application-specific |
| Delete or consolidate small-file trees | Yes | Often depends on the application | Potentially high without an ownership and retention check | Primarily inodes, plus blocks | Any filesystem |
| Remove an obsolete snapshot | Often, after snapshot metadata is updated | No, but filesystem tools may take time | Loss of rollback or backup data | Bytes and metadata | Filesystem-specific, including copy-on-write systems |
| Expand the correct filesystem or move data | Creates capacity rather than deleting existing data | Depends on the storage layout | Low when change-controlled; migration errors remain possible | Bytes and, if the new layout provides them, inodes | Device, volume, and filesystem-specific |
| Change reserved-space settings | May increase ordinary availability but does not remove data | Usually no | Reduced recovery margin if set too aggressively | Bytes available to users | Filesystem-specific; never a first response |
When the filesystem is genuinely full
If df, bounded du, inode checks, open-file checks, journal usage, and storage-layer inspection agree, the data really occupies the available capacity. At that point, cleanup is a capacity decision rather than an accounting mystery.
Make room temporarily
- Move approved archives or backup sets to planned storage.
- Apply application retention policies before removing anything manually.
- Stop nonessential jobs that are generating data while you work.
- Leave enough free space for package managers, logs, database checkpoints, and recovery operations.
Increase capacity deliberately
Expand the volume and filesystem that df -T identified, or migrate data to a larger planned filesystem. The commands differ for local partitions, logical volumes, network filesystems, ext4, XFS, Btrfs, and other types; follow the procedure for the actual device and maintain a verified backup. After the change, verify both df -hT and df -i. A larger block device does not automatically solve an inode design problem, and an inode increase does not create data blocks.
Common mistakes that make the incident worse
- Running
du /without-x: totals from other mounts obscure which filesystem is full. - Deleting a log and expecting immediate relief: the owning process may still hold the descriptor.
- Removing files under
/procor/sys: these are virtual interfaces, not ordinary storage directories. - Deleting database or container files by size: internal files and layers can be required for consistency.
- Killing the first PID shown by
lsof: the process may be critical, and another descriptor may still retain the blocks. - Changing reserved-space settings before identifying the filesystem: the setting and its safety margin vary by filesystem and workload.
- Assuming a reboot is a fix: it closes descriptors but does not remove visible data, snapshots, inode pressure, or bad retention policies.
- Deleting active journal files by hand: use journald’s reporting and supported vacuum behavior instead.
A repeatable checklist for future alerts
- Run
df -hT /pathand record the device, filesystem type, and mount point. - Run
df -i /pathto distinguish inode exhaustion from block exhaustion. - Run
sudo du -xhd1 /mountpoint, then drill into the largest directory. - If the totals diverge, run
sudo lsof +L1and map large entries to services. - Close the owning descriptor through a reload or controlled service restart, then verify
df. - Measure journal usage with
journalctl --disk-usageand apply documented retention or vacuum rules. - Inspect separate mounts, container overlays, snapshots, quotas, and filesystem metadata.
- If accounting reconciles and the margin is still inadequate, expand or migrate the correct filesystem under change control.
The key decision is simple: a high df number is evidence of allocated capacity, not proof that more visible files should be deleted. Match the remedy to the accounting model—close the process for open deleted files, remove approved data for genuine block usage, reduce file counts for inode exhaustion, and use filesystem-specific tools for snapshots, overlays, and metadata.
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.
