- Test filters, indexed sources, exclusions, and service state before rebuilding again.
- Verify database permissions, network access, startup options, and portable profile persistence.
- Use a clean temporary profile to isolate configuration from Windows problems.
- Confirm the Symptom With a Small Safe Test
- Check the Search View Before Blaming the Index
- Confirm the Correct Volumes and Folders Are Indexed
- Check the Everything Service and Account Context
- Check Database Location, Permissions, and Startup Behavior
- Investigate Windows, Network, and Security Boundaries
- Use Diagnostics to Identify What Rebuild Missed
- Run a Clean Temporary Profile Test
- Quick Fix Checklist
- Frequently Asked Questions
When Force Rebuild completes in voidtools Everything but filenames are still missing, stale, duplicated, or associated with the wrong location, rebuilding again is unlikely to help. A rebuild recreates Everything's index from the sources it is currently configured and permitted to read. It does not correct an active search filter, add an omitted volume, repair folder permissions, reconnect an unavailable share, start the Everything service, or remove an exclusion.
The fastest way to solve this symptom is to identify whether the problem is in the search view, index configuration, data source, service, account context, or database storage. The steps below begin with a controlled filename test and then move through the relevant settings. After each change, check whether the test file appears correctly. Once it does, stop changing settings. This prevents a simple problem from turning into several overlapping ones.

Start with free Canva bundles
Browse the freebies page to claim ready-to-use Canva bundles, then get 25% off your first premium bundle after you sign up.
Free to claim. Canva-ready. Instant access.
1. Confirm the Symptom With a Small Safe Test
Before changing the index, determine exactly what Everything is getting wrong. Choose a local folder you control, such as a folder inside your Windows user profile. Avoid using a protected system directory or an unreliable network share for the first test.
- Create an empty text file with a distinctive name, such as
everything-index-test-4827.txt. - Open Everything and clear the search box completely.
- Select the Everything filter rather than Audio, Documents, Folders, or another specialized filter.
- Search for the complete distinctive filename.
- Rename the file to
everything-index-test-5931.txtand search again. - Delete the file and verify that the old result disappears.
This test distinguishes several failure types. If the original file never appears, its volume or folder might not be indexed. If it appears only after Force Rebuild, real-time change monitoring might not be working. If both the old and new names remain, Everything may be reading stale data, indexing the same source twice, or receiving updates incorrectly. If the test works but another location does not, focus on that location rather than the whole application.
Success means the new file appears, the rename is reflected, and the deleted name disappears without requiring repeated rebuilds. If all three actions work, the core local index is functioning. Stop rebuilding and investigate the specific missing drive, folder, share, search option, or filter.
2. Check the Search View Before Blaming the Index
A correct index can produce apparently wrong results when the active search view hides matching entries. Force Rebuild does not reset search options, filters, result omissions, or command-line search parameters.
2.1 Clear the active filter and search modifiers
Set the active filter to Everything and remove all text from the search box. Then type only the distinctive test filename. Check the Search menu for options such as Match Case, Match Whole Word, Match Path, Match Diacritics, and regular-expression mode. A search written for one mode can behave very differently in another.
Also look for search syntax left in the query. Operators, macros, path restrictions, extension searches, negation, and regular expressions can intentionally remove results. If the status bar shows a small result set while you expect the entire index, the current query or filter is probably narrowing the view.
Success means the missing item appears after selecting the Everything filter or simplifying the query. If it does, the index was not broken. Restore search options one at a time until you identify the setting or expression that hid the result.
2.2 Check exclusions and result omissions
Open Everything's Options and review the exclusion settings. Confirm that the affected path, filename pattern, file attribute, or volume is not excluded. Pay particular attention to exclusions created long ago for temporary folders, backups, hidden files, system files, or removable drives.
Everything can also omit results from the current view without removing them from every underlying source. Review any omit-results feature or related search command you use. If you launch Everything through a shortcut, script, launcher, or portable-app menu, inspect that launch method for command-line options that change what is displayed.
Success means removing or correcting the relevant exclusion makes the test item appear after the index updates. Do not remove every exclusion indiscriminately. Keep intentional exclusions and modify only the rule that matches the affected location.
3. Confirm the Correct Volumes and Folders Are Indexed
Force Rebuild processes configured sources. If the required source is absent, unavailable, or configured in the wrong indexing mode, a successful rebuild can still produce incomplete results.
3.1 Verify local NTFS and ReFS volumes
In the Indexes section of Options, inspect the configured NTFS and ReFS volumes where applicable. Confirm that the affected drive is listed, included in the database, and currently accessible. Drive letters can change after reconnecting external disks, cloning drives, modifying partitions, or attaching removable storage.
Check the volume identity and path rather than assuming that a familiar drive letter still refers to the same disk. If a drive was replaced, an old offline volume entry can coexist with the new one and create missing or confusing results.
Success means the expected volume is present, enabled, and its current files can be found. If the local test works but files on one drive remain absent, continue troubleshooting only that drive.
3.2 Verify folder indexes for shares and nonstandard file systems
Network shares, mapped drives, NAS locations, and file systems that cannot be indexed through the local NTFS mechanism generally need folder indexing or another supported connection method. Review the Folders page in Options and verify that the exact path is included.
A mapped drive is tied to a Windows logon session and account context. A service or elevated process may not see the same mapped drive letters visible in a normal File Explorer window. When practical, test the share using its UNC path, such as \\server\share, instead of relying only on a mapped letter.
Folder indexes are scanned according to their configuration. They may not update with the same immediacy as a directly monitored local NTFS volume. Confirm that the folder exists, can be opened under the account running Everything, and has completed its scheduled or manual rescan.
Success means the share or folder can be browsed under the relevant account and its distinctive test file appears after a folder rescan. Once it works, stop changing local NTFS settings because they are not responsible for that source.
3.3 Look for duplicate indexing paths
Duplicate results can occur when the same content is indexed through more than one route, such as both a mapped drive and its UNC path, or both a local volume and an added folder beneath that volume. Compare the paths in the results and the configured sources.
Remove only the redundant source after confirming which path applications and users need. Success means each file appears once under the intended path without removing access to unique content.
4. Check the Everything Service and Account Context
The Everything service helps the desktop application access local file system information without requiring the interface itself to run as administrator. A stopped, missing, or mismatched service can prevent correct indexing even though the application opens normally.
4.1 Verify service state
Open the service settings in Everything or use the Windows Services management console to check the Everything service. Confirm that it is installed when your configuration expects it, is running, and is not repeatedly stopping. Restarting the service once is a reasonable diagnostic step. Repeatedly restarting it without checking the result is not.
If Everything is configured to use the service, confirm that the application can connect to it. Avoid running multiple copies with different configurations unless that is intentional. A portable copy, installed copy, and background service can otherwise make it difficult to tell which instance owns the visible index.
Success means the service remains running and a newly created local file appears promptly. If the service will not start, check Windows event information, installation permissions, and security-software logs rather than performing another Force Rebuild.
4.2 Compare elevated and standard-user behavior
Run the controlled test in your normal account first. If Everything behaves differently when launched as administrator, account context is relevant. Elevated processes can have different access to mapped drives and user-specific settings. Conversely, a standard account may not be able to write to a database location chosen under a protected directory.
Use one consistent launch method while diagnosing the problem. Check shortcut properties, startup entries, scheduled tasks, and portable launchers for options that select another configuration file, instance name, database path, or server.
Success means the same expected index opens each time under the intended account. Once you identify a shortcut or startup command loading another profile, correct that launch method instead of rebuilding both profiles.
5. Check Database Location, Permissions, and Startup Behavior
Everything needs to read and update its database and configuration. A Force Rebuild may appear to run during the session, but changes can be lost or an older database can return at the next launch if the selected location is not writable.
5.1 Test database write access safely
Find the configured database location in Options. Confirm that the current user or relevant process can create and modify files there. Common problem locations include protected program directories, read-only portable media, disconnected synchronized folders, network paths with intermittent access, and folders controlled by another account.
Do not begin by deleting the database. First close Everything normally and verify that its files can be written in the configured directory. If needed, select a temporary local folder owned by your user for a diagnostic profile. Preserve the existing database and settings until the cause is known.
Success means the rebuilt index survives a normal exit and relaunch, and its modification time changes when expected. If results are correct until restart and then become stale again, database persistence or profile selection is a stronger suspect than scanning.
5.2 Inspect startup and portable configuration
Portable users should confirm where Everything stores its INI file and database. A launcher may copy an old configuration into place at startup or discard changes at shutdown. Likewise, two shortcuts may start different instances or point to different configuration locations.
Temporarily start Everything directly from the intended executable without the portable-app wrapper. Do not add special command-line switches. If the direct launch works, inspect the wrapper's persistence, paths, and parameters.
Success means reopening Everything returns to the same options, source list, and current results. At that point, stop rebuilding and repair the startup or portable configuration process.
6. Investigate Windows, Network, and Security Boundaries
When only a NAS, share, mapped drive, or protected directory is wrong, Windows access is often the underlying cause. Force Rebuild cannot index content that the active account cannot enumerate.
6.1 Confirm permissions and share availability
Using the same Windows account and elevation level as Everything, open the affected path in File Explorer. Create a harmless test file only if you have permission and doing so is appropriate. Confirm that folder listings are complete and that credentials have not expired.
For a NAS, check whether the host is awake, the share name is unchanged, and the UNC path resolves. A mapped drive displayed in Explorer does not guarantee that a service, scheduled task, or elevated process can access it.
Success means Everything can scan the path after the same account can enumerate it reliably. If Explorer itself cannot list the directory in that context, fix connectivity or permissions before changing Everything again.
6.2 Review firewall and antivirus events carefully
Security software can block service communication, database writes, executable launches, or network access. Review recent protection history and firewall logs for events involving Everything, its service, its database directory, or the remote share.
Do not permanently disable antivirus or firewall protection. If testing is necessary, use a narrow, temporary, documented rule and restore protection immediately. For network-server features, restrict access to trusted interfaces and networks. Never expose an Everything search server directly to the public internet without an appropriate secured architecture.
Success means a specific approved rule allows the required local or trusted-network operation while security controls remain active. If there is no relevant security event, do not create broad exclusions merely as a guess.

7. Use Diagnostics to Identify What Rebuild Missed
Diagnostics are most useful after the simple filter, source, service, and permission checks. The goal is to observe where the expected item disappears, not to collect large logs without a test case.
7.1 Read the status bar and Index Journal
The status bar can reveal whether Everything is still loading, rebuilding, scanning, or showing only a filtered subset of its database. Wait for the current operation to finish before judging results.
Where available in your installation, use the Index Journal or related index information to check whether file changes are being recorded. Create, rename, and delete the distinctive test file while observing whether those changes reach the index. If an initial rebuild is correct but later changes are absent, update monitoring is the likely issue.
Success means each test action produces the expected index change. If rebuild results are correct but subsequent changes are not, focus on monitoring, service state, folder rescan behavior, or source availability.
7.2 Use narrow search syntax tests
Search for the exact filename first, then test its parent path separately. If a name search works but a path-constrained search does not, Match Path or the path expression may be wrong. If the parent folder is missing entirely, the source or an exclusion is more likely responsible.
Avoid complicated regular expressions during diagnosis. Begin with literal names that contain unusual digits, then add one operator at a time. This makes it clear which part of the query changes the result.
7.3 Capture a short debug log
If the cause remains unclear, enable Everything's supported debugging or logging facilities, reproduce one create, rename, or rescan failure, and then stop logging. Record the time, source path, account, active filter, and expected filename.
Review the log for failed access, unavailable volumes, database write errors, service connection problems, or unexpected configuration paths. Remove or redact sensitive filenames and server details before sharing logs with support personnel.
Success means the log identifies a concrete failed operation or confirms that the item entered the index and was later hidden by the search view. Act on that evidence rather than making several unrelated changes.
8. Run a Clean Temporary Profile Test
A brand-new temporary profile is one of the safest ways to separate configuration problems from Windows or storage problems. It should be a test, not an immediate replacement for your working configuration.
- Exit existing Everything instances normally.
- Preserve the current configuration and database rather than deleting them.
- Start a separate temporary instance or clean portable copy according to the supported instance and configuration methods.
- Use a new writable local directory for its configuration and database.
- Do not import filters, exclusions, command-line options, servers, or old settings.
- Index only one safe local source needed for the distinctive filename test.
- Create, rename, and delete the test file.
If the clean profile works, Windows and the local source are probably usable. Compare the old and new profiles for filters, exclusions, indexed sources, service settings, database paths, instance options, and startup arguments. Reintroduce required settings individually.
If the clean profile fails in the same way, the cause is more likely the service, source availability, file system, account permissions, or security boundary. There is little value in repeatedly resetting application preferences at that point.
Success means the clean profile tracks all three test actions. Stop making broad changes once you have isolated the fault to either the original configuration or the external environment.
9. Quick Fix Checklist
- Clear the query and select the Everything filter.
- Disable only unintended search modes and test with a literal unique filename.
- Confirm the affected volume or folder is included in the index.
- Check that exclusions and result omissions do not match the missing path.
- Verify folder indexes and rescans for NAS paths, shares, and nonstandard file systems.
- Use the UNC path when mapped-drive visibility differs by account context.
- Confirm the Everything service is running when the configuration depends on it.
- Use one executable, profile, instance, and launch method during diagnosis.
- Verify that the database directory is writable and persists across restart.
- Check startup shortcuts and portable wrappers for old configuration or command-line options.
- Review firewall and antivirus logs without permanently disabling protection.
- Run a clean temporary profile test before replacing existing settings.
After each item, repeat the same create, rename, and delete test. Stop when the behavior becomes correct. A working result is more valuable than completing every checklist item, and unnecessary later changes can hide the original cause.
10. Frequently Asked Questions
10.1 Why does Force Rebuild finish but Everything results remain missing?
Force Rebuild reconstructs the database from sources Everything can currently access and is configured to include. Results can remain missing when a filter hides them, the relevant volume or folder is not included, an exclusion matches the path, a network source is unavailable, or the running account lacks access.
10.2 Is Everything the same as Windows Search?
No. voidtools Everything is a separate filename-search application with its own index, configuration, service, filters, and folder-indexing features. Rebuilding Everything does not rebuild the Windows Search index, and Windows Search troubleshooting steps generally do not repair an Everything configuration problem.
10.3 Why does Everything work after rebuilding but become stale later?
This pattern suggests that the initial scan succeeds but later file changes are not being monitored or rescanned. Check service continuity, source availability, folder-index schedules, file system support, and whether the database can be saved. Use the distinctive rename test to verify real-time updates.
10.4 Why are mapped-drive results missing when the drive works in File Explorer?
Mapped drives belong to a particular user session and may not be visible to an elevated application, service, or scheduled task. Test the UNC path under the same account context as Everything, then configure an appropriate folder index or supported server arrangement.
10.5 Should I delete the Everything database?
Not as the first step. Force Rebuild already recreates index data, and deleting files can obscure a profile, permission, source, or startup problem. First verify filters, indexed sources, exclusions, service state, write permissions, and the behavior of a clean temporary profile. Preserve the original database until diagnosis is complete.
10.6 When should I stop troubleshooting?
Stop when your controlled test file appears, updates after rename, disappears after deletion, and remains correct after a normal restart. If only one remote or protected location still fails, narrow the remaining work to that source. Do not keep resetting unrelated settings after the main index behaves correctly.