About this article: This article is reviewed against the cited sources. This is source research, not lab validation. Check the stated versions and your own requirements before making changes.
You try to create a new user, group or computer account and get:
Windows can't create the object because the Directory Service was unable to allocate a relative identifier.
Every user, group and computer in a domain gets a security identifier (SID), and the last part of it is a relative identifier (RID). Domain controllers hand out RIDs from blocks they get from one domain controller, the RID master. If a domain controller has used up its block and cannot get a new one, account creation on that domain controller stops (Microsoft Learn, Managing RID issuance).
There are two very different versions of this fault. It can be local: one domain controller cannot reach the RID master (Microsoft notes that many RID issues affect only one domain controller). Or, much more seriously, the whole domain is close to running out of RIDs. The checks below tell you which one you have.
What you see
- Account creation fails with the message above, in Active Directory Users and Computers, PowerShell or a provisioning tool.
- It may fail on one domain controller and work on another.
- The System event log, source Directory-Services-SAM, may hold RID events. Microsoft lists these (Microsoft Learn, Managing RID issuance):
| Event | Level | Meaning |
|---|---|---|
| 16642, 16643, 16644, 16645 | — | Older ("legacy") RID pool request problems on a domain controller |
| 16653 | Warning | The RID Block Size registry value is above the maximum of 15,000; 15,000 is used instead |
| 16654 | Informational | A domain controller's RID pool was thrown away, for example after a restore from backup or a virtual machine snapshot |
| 16655 | Informational | The domain's total RID space was increased by an administrator |
| 16656 | Warning | The domain has used a large share of its RIDs and is nearing the protection ceiling |
| 16657 | Error | The ceiling was reached: the RID master has stopped issuing RID pools |
| 16658 | Warning | Periodic count of remaining RIDs; only appears after about 100 million RIDs have been used |
Why it happens
One domain controller cannot get a new pool. A domain controller normally asks for 500 RIDs at a time. When its pool runs out it must contact the RID master. If the RID master is offline, unreachable, or replication between them is broken, the request fails and that domain controller cannot create accounts. Other domain controllers with RIDs left keep working, which is why the fault can look random.
The RID master was lost and not replaced. If the server holding the RID master role failed or was removed without the role being moved, no domain controller can get new pools once theirs run out (Microsoft Learn, transfer or seize roles).
The domain is running low on RIDs. By default a domain has about one billion RIDs (2³⁰ − 1 = 1,073,741,823). Microsoft has seen domains use them up through scripts or provisioning tools that bulk-created accounts by mistake, many unused groups, many domain controllers demoted, restored or cleaned up, repeated forest recoveries, repeated pool invalidation, or a RID Block Size value set too high. When only 10% remain, the RID master stops issuing pools on purpose (event 16657) so someone has to look (Microsoft Learn, Managing RID issuance).
Also check the simple causes first. Microsoft's own order is: does the person have permission to create the object, and is a domain controller available at all?
How to prove the cause
All of these are read-only.
1. Rule out permissions and availability. Look for "access denied" or domain controller availability messages in the full error output.
2. Read the System log on the domain controller that failed and on the RID master, filtered to source Directory-Services-SAM. Events 16642–16645 point to a pool request problem; 16656 or 16657 point to the domain-wide ceiling. Microsoft says that if none of the newer events appear, treat it as a classic pool request problem.
3. Find the RID master. In Active Directory Users and Computers, right-click the domain, choose Operations Masters and open the RID tab (Microsoft Learn, view and transfer FSMO roles). Check that the named server still exists and is running.
4. Test the RID master and the remaining pool.
Dcdiag.exe /test:ridmanager /v
This checks that the RID master can be reached. To see the domain's total RID space:
Dcdiag.exe /TEST:RidManager /v | find /i "Available RID Pool for the Domain"
The default total is 1,073,741,823; after an unlock it is 2,147,483,647.
5. Check replication between the failing domain controller and the RID master with Repadmin. Microsoft suggests a network capture at both ends if these tests do not settle it.
The fix
If the RID master is reachable but replication or the network is broken, fix that first. Once the failing domain controller can talk to the RID master, it gets a new pool and account creation works again.
If the RID master is healthy but you need to move the role (for example before maintenance), transfer it. In Active Directory Users and Computers, connect to the domain controller that will take the role, open Operations Masters, select the RID tab and click Change (Microsoft Learn, view and transfer FSMO roles). Ntdsutil can do the same with transfer rid master.
If the RID master is gone for good, seize the role onto a healthy domain controller in the same domain with the most up-to-date copy of the domain (Microsoft Learn, transfer or seize roles). As a Domain Admin:
ntdsutil
roles
connections
connect to server <new RID master name>
q
seize rid master
q
q
Seizing deliberately skips ahead in the RID sequence to avoid duplicate SIDs: Ntdsutil skips 10,000 RIDs and Move-ADDirectoryServerOperationMasterRole skips 30,000. Seize only when you are sure the old RID master will not return. Afterwards the old server must never come back as a domain controller from its old disk or a backup: rebuild or force-demote it, and clean up its metadata on another domain controller first.
If event 16657 says the ceiling was reached, do not just switch issuing back on. Microsoft's event text says to find and fix whatever is using RIDs at an abnormal rate first, because a domain that runs out completely can never create accounts again without a domain migration or forest recovery. Once the cause is fixed, an administrator re-enables issuing on the RID master by setting msDS-RIDPoolAllocationEnabled to TRUE on CN=RID Manager$,CN=System,DC=<domain> with LDP, following Microsoft's step-by-step procedure. Microsoft also documents a one-time "unlock" that doubles the RID space; it cannot be undone, older domain controllers cannot issue RIDs afterwards, and Microsoft says not to use it pre-emptively (Microsoft Learn, Managing RID issuance).
How to confirm the fix
- Creating a test account works on the domain controller that failed.
Dcdiag.exe /test:ridmanager /vpasses.- No new 16642–16645 or 16657 events appear in the System log.
What to watch afterwards
- Directory-Services-SAM events on every domain controller. Microsoft recommends that monitoring systems alert on 16653 to 16658; any 16656, 16657 or 16658 is a serious sign.
- Provisioning scripts and tools. A failed account creation still uses up a RID, for example creating a user with a password that does not meet complexity rules. A script that retries in a loop can burn through RIDs.
- Event 16654 you did not expect. It means a domain controller's pool was discarded, for example by restoring a virtual machine snapshot. Find out who did it and why.
- Where the RID master lives. Keep it on a reliable, well-connected domain controller, and move it before you retire or rebuild that server.
Read next
Related articles you may want next.
- Backup & RecoveryActive Directory Forest Recovery: Authoritative Restore, RID Pool and DC OrderWhich domain controller to restore first, what authoritative means for the directory and SYSVOL, why the RID pool is raised, and health checks.Read article
- TroubleshootingAD replication error 1722 "The RPC server is unavailable": find the layer that is really failingActive Directory replication fails with 1722 (0x6ba). Prove whether name resolution, a blocked port, signing or the network is at fault, fix that layer and confirm replication has caught up.Read article
- Backup & RecoveryRecovering deleted Active Directory objects: the Recycle Bin first, backups secondUnderstand Active Directory object recovery, Recycle Bin prerequisites, parent order, retention limits and backup-based recovery decisions.Read article
Risks and limits
- Seizing the RID master while the old one is still alive, or bringing the old one back after a seizure, risks two RID masters and duplicate RID pools. Microsoft lists the RID master among the roles that cause conflicts when duplicated.
- Do not raise the RID Block Size registry value to "fix" this. It speeds up RID use; above 15,000 it is ignored and logs event 16653 at every restart.
- The ceiling override and the RID space unlock are domain-wide changes made with LDP against the RID master. Follow Microsoft's page exactly, and have a tested backup of Active Directory first.
Sources
- Microsoft Learn, "Managing RID Issuance", applies to Windows Server 2016–2025, last updated 12 May 2025: https://learn.microsoft.com/en-us/windows-server/identity/ad-ds/manage/managing-rid-issuance
- Microsoft Learn, "Transfer or seize Operation Master roles in Active Directory Domain Services" (KB 255504), last updated 7 April 2025: https://learn.microsoft.com/en-us/troubleshoot/windows-server/active-directory/transfer-or-seize-operation-master-roles-in-ad-ds
- Microsoft Learn, "How to view and transfer FSMO roles" (KB 324801), last updated 12 February 2026: https://learn.microsoft.com/en-us/troubleshoot/windows-server/active-directory/view-transfer-fsmo-roles
Next step
When a RID error comes from one domain controller losing touch with the RID master, the real work is fixing that path without making things worse by seizing roles too early. NetLeverage has looked after business servers since 2002. If accounts will not create and you want a careful check of your domain controllers, talk to us through managed support with the exact error text and any Directory-Services-SAM event IDs.
This article was originally published on .