Fix it now
0x80244019 is WU_E_PT_HTTP_STATUS_NOT_FOUND, the same as HTTP 404: the server cannot find the requested URI. 0x80244022 is WU_E_PT_HTTP_STATUS_SERVICE_UNAVAIL, the same as HTTP 503. Both are server-side, and reimaging clients will not help.
wuauclt.exe /resetauthorization /detectnow
- Start on a failing client, in a browser, with Microsoft’s three documented connectivity tests. Substitute your own server name and port:
http://<server>:8530/Selfupdate/wuident.cab,http://<server>:8530/ClientWebService/wusserverversion.xmlandhttp://<server>:8530/SimpleAuthWebService/SimpleAuth.asmx. - If all three fail, check the WSUS Administration site and the WsusPool application pool in IIS Manager. Microsoft states that when you get an HTTP 503 or a connection error to the WSUS Administration website, the most likely cause is that WsusPool is stopped.
- Start the pool and watch whether it stops again within the hour. If it does, the memory limit is the problem, not the start.
- Raise it: select WsusPool, choose Advanced Settings, and set Private Memory Limit (KB) to 4000000. The default is 1843200 KB, and Microsoft notes it may be necessary to go to 8000000 KB or higher depending on the environment.
- If only some of the three URLs fail, the virtual directory behind that one is the problem rather than the whole site.
- Once the server answers, force one client to rescan with
wuauclt.exe /resetauthorization /detectnowrather than waiting for its next cycle.
If the tests succeed from the client, review the IIS logs on the WSUS server to confirm the HTTP errors are actually coming from WSUS. If they are not, the problem is an intermediate firewall or proxy.
If clients report in with a current last-contact time, you are done. The next section covers why a 404 and a 503 point at different layers.
Why it happens
A WSUS client does not talk to a database or a file share. It makes SOAP calls to a set of web services hosted in IIS on the WSUS server: a client web service for scanning, a simple authentication service, a reporting service and a self-update path for the agent itself. Content is served separately from a content virtual directory. Each of those is a URL, and each can independently be missing, stopped or wrongly bound.
That is why a 404 is more informative than it looks. It does not mean the server is down; the machine answered. It means the specific virtual directory the client asked for is not registered in IIS, or is bound to a different port than the one the client is using. Since Windows Server 2012, WSUS uses 8530 for HTTP and 8531 for HTTPS by default. If clients are configured for port 80 because someone copied an old policy, every one of them gets a 404 from the default web site while WSUS sits perfectly healthy on 8530.
A 503 is the opposite: the URL is right and IIS refuses to run the application behind it. Microsoft’s own guidance is that a stopped WsusPool is the most likely cause of an HTTP 503 or a connection error to the WSUS Administration website, and the usual reason the pool stops is the private memory limit. The default of 1843200 KB was chosen for much smaller deployments than most estates now run, and a server carrying years of metadata exceeds it during a sync and takes the pool down with it.
Be careful not to turn that into a definition. Microsoft describes 0x80244022 more broadly than a stopped pool: unable to connect to the configured update source, with network troubleshooting needed to resolve the connectivity issue. The stopped pool is the first thing to check, not the meaning of the code. 0x80244018 is WU_E_PT_HTTP_STATUS_FORBIDDEN, HTTP 403: the server understood the request and declined it, which is authentication or authorisation rather than availability. 0x80244021 is WU_E_PT_HTTP_STATUS_BAD_GATEWAY, HTTP 502, which by definition comes from a gateway or proxy in front of WSUS rather than from WSUS itself.
The application pool has stopped
You have this one if 0x80244022 across the estate, and WsusPool shows as stopped in IIS Manager.
- Start WsusPool and confirm clients recover. If it stops again within the hour, the limit is the problem.
- In IIS Manager select Application Pools, right-click WsusPool and choose Advanced Settings.
- Set Private Memory Limit (KB) to 4000000, up from the default 1843200. Microsoft notes 8000000 KB or higher may be necessary depending on the environment.
- Review the Recycling settings on the same pool and remove any aggressive scheduled recycle that coincides with your sync window.
Where WSUS is a Configuration Manager software update point, Microsoft’s best practices give a fuller set: Queue Length 2000, Idle Time-out 0, Ping Enabled False, Private Memory Limit 0, Regular Time Interval 0.
Clients are pointed at the wrong port or URL
You have this one if 0x80244019 on every client, while the WSUS console and the server itself look completely healthy.
- On a client, check
WUServerandWUStatusServerunderHKLM\SOFTWARE\Policies\Microsoft\Windows\WindowsUpdate. - Confirm they include the port WSUS actually uses. Since Windows Server 2012 that is 8530 for HTTP and 8531 for HTTPS.
- Correct the Group Policy object rather than the client registry, then run
gpupdate /forceon the client. - Test the corrected URL from the client’s browser before rescanning.
A virtual directory is missing or broken in IIS
You have this one if One of Microsoft’s three test URLs returns 404 while the others answer.
- In IIS Manager, expand the WSUS Administration site and confirm ClientWebService, SimpleAuthWebService, ServerSyncWebService, ReportingWebService, SelfUpdate and Content are all present.
- If any are missing, re-run the post-installation step:
WsusUtil.exe postinstallfrom the WSUS tools folder, supplying the content directory if you use one. - Confirm the site bindings still list the ports you expect.
- Retest the three URLs from a client.
The server is refusing rather than failing to find
You have this one if 0x80244018 or 0x80244021 instead of a clean 404 or 503.
- Check the client certificate setting on the WSUS website and its virtual directories. Microsoft names virtual directories configured to Accept or Require client certificates as a cause of these scan failures.
- Check the authentication settings. A hardening baseline that removes anonymous access from the client web services breaks every client.
- If a load balancer or reverse proxy sits in front of WSUS, check it is not returning its own error pages. A 502 comes from the device in front, not from WSUS.
- Confirm any request filtering or URL rewriting rules added to IIS are not catching the WSUS paths.
The server is overloaded during synchronisation
You have this one if Failures cluster at the same time each day and clear on their own.
- Move the synchronisation schedule away from the period when most clients scan.
- Run a WSUS cleanup to decline superseded updates and remove unused content, which reduces the metadata the pool has to hold.
- Check disk performance on the volume holding the WSUS database, which is usually the real bottleneck.
- Stagger client scan times through policy rather than letting an estate scan at once.
Full reference
Matching the code to the layer
| Code | HTTP | Published name and meaning |
|---|---|---|
| 0x80244019 | 404 | WU_E_PT_HTTP_STATUS_NOT_FOUND: the server cannot find the requested URI |
| 0x80244022 | 503 | WU_E_PT_HTTP_STATUS_SERVICE_UNAVAIL: the service is temporarily overloaded. Microsoft’s broader mitigation is that the client is unable to connect to the configured update source |
| 0x80244018 | 403 | WU_E_PT_HTTP_STATUS_FORBIDDEN: the server understood the request but declined to fulfil it |
| 0x80244021 | 502 | WU_E_PT_HTTP_STATUS_BAD_GATEWAY: a gateway or proxy received an invalid response from the upstream server |
Microsoft’s documented connectivity tests
| URL | What a success proves |
|---|---|
http://<server>:8530/Selfupdate/wuident.cab |
The self-update path is present and serving |
http://<server>:8530/ClientWebService/wusserverversion.xml |
The client web service virtual directory is present and serving |
http://<server>:8530/SimpleAuthWebService/SimpleAuth.asmx |
The simple authentication service is present and serving |
Run all three from a failing client rather than from the server. If they succeed from the client and clients still fail, Microsoft’s advice is to review the IIS logs on the WSUS server to confirm the HTTP errors are actually being returned by WSUS. If WSUS is not returning them, the error is being generated by an intermediate firewall or proxy and the WSUS server is not your problem.
Application pool settings, with the numbers
| Setting | Default | Documented value |
|---|---|---|
| Private Memory Limit (KB) | 1843200 | 4000000, or 8000000 and higher depending on the environment |
| Queue Length | 1000 | 2000, where WSUS is a Configuration Manager software update point |
| Idle Time-out (minutes) | 20 | 0, in the same scenario |
| Ping Enabled | True | False, in the same scenario |
| Regular Time Interval (minutes) | 1740 | 0, to prevent a recycle, in the same scenario |
The two sets are for different situations. Raising Private Memory Limit to 4000000 KB is Microsoft’s general remedy for a pool that keeps stopping. The fuller set, including Private Memory Limit 0 for unlimited, is published for a WSUS server acting as a Configuration Manager software update point. Removing the cap trades a hard failure for the risk of unbounded growth, so give the server the memory to back it.
Capacity, before you blame the software
Microsoft states that WSUS can support 100,000 clients per server, and 150,000 when used with Configuration Manager, and says in the same breath not to approach that limit. The recommended approach at scale is two to four servers sharing the same SQL Server database rather than one server run close to its ceiling. If your estate is large and your pool keeps falling over during sync, that is a capacity conversation rather than a tuning one.
Ports and what has to be open
| Connection | HTTP | HTTPS |
|---|---|---|
| Client to WSUS | 80 or 8530 | 443 or 8531 |
| WSUS to an upstream WSUS server | 80 or 8530 | 443 or 8531 |
The pairing is fixed: if the HTTP port is 80 the HTTPS port must be 443, and if the HTTP port is anything else the HTTPS port follows the same scheme, as with 8530 and 8531. Where HTTPS is configured, the HTTP port still has to remain open, because some data such as update licence terms is served unencrypted.
Client-side, once the server is healthy
- Clients pick a healthy server up at their next scan. You do not usually need to touch them.
- To confirm immediately rather than waiting, Microsoft still publishes
wuauclt.exe /resetauthorization /detectnowas a client-side step in current troubleshooting guidance. - Read
WUAHandler.logandWindowsUpdate.logon a Configuration Manager client for what the agent reported. - Check the client is not also failing name resolution, which produces a different code entirely and is a different article.
- If one client is stuck while the rest recover, treat it as a client problem rather than continuing to change the server.
Every code this article covers
| Code | What it points at | Source |
|---|---|---|
0x80244019 |
WU_E_PT_HTTP_STATUS_NOT_FOUND: the same as HTTP 404, the server cannot find the requested URI | Microsoft Learn |
0x80244022 |
WU_E_PT_HTTP_STATUS_SERVICE_UNAVAIL: the same as HTTP 503, the service is temporarily overloaded. Microsoft’s mitigation is broader than a stopped pool: the client is unable to connect to the configured update source | Microsoft Learn |
0x80244018 |
WU_E_PT_HTTP_STATUS_FORBIDDEN: the same as HTTP 403, the server understood the request but declined to fulfil it | Microsoft Learn |
0x80244021 |
WU_E_PT_HTTP_STATUS_BAD_GATEWAY: the same as HTTP 502, a gateway or proxy received an invalid response from the upstream server | Microsoft Learn |
Confirm the fix worked
- Open all three of Microsoft’s test URLs from a client browser and confirm each returns content rather than an error.
- In IIS Manager, confirm WsusPool is started and stays started through a full synchronisation.
- Trigger a scan on one client and confirm it reports into the WSUS console with a current last-contact time.
- Check the IIS logs on the server and confirm the client requests are returning 200 rather than 404 or 503.
- Confirm a second client recovers on its own at its next scan, without being touched.
Questions people ask about this
Does a 503 always mean the application pool stopped?
It is the first thing to check, and Microsoft names it as the most likely cause of an HTTP 503 against the WSUS Administration website. But Microsoft’s own description of 0x80244022 is broader: the client is unable to connect to the configured update source, which can also be a network or proxy problem between the client and the server.
Should I set Private Memory Limit to 0?
That value comes from Microsoft’s best practices for a WSUS server acting as a Configuration Manager software update point, alongside Queue Length 2000, Idle Time-out 0, Ping Enabled False and Regular Time Interval 0. For a pool that simply keeps stopping, the documented remedy is 4000000 KB, going to 8000000 KB or higher if the environment needs it. Removing the cap entirely means giving the server enough memory to back it.
Does any of this need a licence or a paid product?
No. WSUS is a role included with Windows Server, and every fix here is a configuration change on a server you already own. Whether the server itself is correctly licensed is a separate question from these codes.
How many clients can one WSUS server take?
Microsoft publishes 100,000 clients per server, and 150,000 when used with Configuration Manager, while saying not to approach those limits. At scale the recommended arrangement is two to four servers sharing the same SQL Server database.
Do I need to touch the clients at all?
Usually not. Once the server answers, clients pick it up at their next scan. You can force one with wuauclt.exe /resetauthorization /detectnow if you want to confirm the fix immediately rather than waiting.
