Running Bitwarden?

Bitwarden self-hosted or SCIM unhappy? We bring it back.

Bitwarden is a well-liked password manager that becomes a real operations job when you choose to self-host it, hook it into SCIM or run the newer Unified container in a shape it does not quite like. We pick up the sync failures, the emergency-access requests that never complete, and the server install that drifted away from the official charts.

  • Since 2015Fixing technology
  • FreeEstimates
  • US teamReal engineers

bitwarden-selfhost.log

still failing

What it is

Bitwarden at a glance

Bitwarden is an open-source password manager with a free hosted tier, a paid business tier and a self-hosted option that enterprises choose when they want the data to stay on their own infrastructure. It was founded in 2016 and has grown steadily as the main open-source alternative to 1Password.

The self-hosted deployment runs as a set of Docker containers with Microsoft SQL Server or PostgreSQL behind them. The newer Unified image consolidates the stack into one container for smaller teams, and the older multi-container install remains the supported path for larger deployments. Both work well once dialed in; both have their own particular ways to break during an upgrade or a volume migration.

Type
Open-source password and secrets manager, with self-hosted option
First released
2016
License
GPLv3 for clients and server, with a commercial enterprise license
Maintained by
Bitwarden, Inc.
Typically used for
Personal and business password vaults, self-hosted deployments, SCIM-driven team management, developer secrets via CLI

The honest version

Bitwarden pros and cons

No technology is all upside. This is where Bitwarden earns its place and where it tends to cause trouble.

Where Bitwarden is strong

  • Open source, auditable, inspectableThe clients, the server, the CLI and the directory connector are all in public repositories, which is a real benefit when a security review asks hard questions.
  • Self-hosting is a credible optionYou can keep every encrypted vault on your own infrastructure, which matters for regulated industries and for teams that do not want a third party in the path.
  • Reasonable price per seatThe business tier is meaningfully cheaper per seat than the common alternatives, which adds up quickly across a few hundred users.
  • Good CLI and SDKThe bw command and the newer SDK make it usable for scripted deploys, CI and developer-secret flows.
  • Directory Connector and SCIMUsers and groups can be synced from Active Directory, Okta, Azure or Google with an onshore connector or via SCIM.
  • Emergency access and fingerprintA trusted contact can request access after a wait, and the fingerprint phrase gives clients a human-checkable identity for shared vaults.

Where Bitwarden bites

  • Self-hosting is a real operations jobContainers, SQL backups, TLS renewals, SMTP for invitations and push notifications all have to work. Nothing about the install is self-maintaining.
  • The Unified image hides sharp edgesIt is marketed for small teams and will absolutely run a larger one until a migration or a backup restore brings out limitations that the multi-container setup does not share.
  • SCIM is less mature than it looksPush provisioning from Okta or Azure works, mostly, until an IdP sends an operation the server rejects and the admin finds out through a user complaint.
  • Attachments and icons storage can fill upVolumes are undersized by default, nobody watches them, and the day they fill the API returns 500 on sync.
  • Emergency access needs a tested runbookThe flow works, but only if the trusted contact knows the steps and the wait time has not been set so long that the request expires.
  • Upgrade path can biteContainer versions, database schema, and client minimum versions drift. Skipping a few versions and jumping to the latest is not supported in the way admins hope.

Sound familiar?

Common Bitwarden problems we fix

The Bitwarden problems that bring people to us, and what you have once each one is fixed.

500 on /sync

Clients fail to sync against a self-hosted server

A database connection is exhausted, the attachments volume is full, or the API container ran out of memory overnight. Clients fail to pull vault updates and nobody can log in.

Server syncing again under load

Kdf iterations mismatch

Vaults fail to decrypt after a master password change

The KDF settings on the account and the client fall out of sync, often after a protocol upgrade. The vault opens on one device and refuses on another.

KDF consistent, vault decrypts

SCIM: externalId exists

SCIM push creates duplicate users

The IdP is posting users with an externalId that does not match what the server already stored, or provisioning was flipped off and on and the mapping was lost.

Idempotent SCIM, one user per person

emergency-access expired

Emergency access requests never complete

The wait time was set to days, nobody granted the request in time, or the trusted contact's account itself was deprovisioned before they could act.

Emergency access tested, reachable

attachments volume 94% full

Storage volumes are running out

The default allocation for attachments and icons is small; a few years of use fills it, and when it hits full the API starts returning server errors.

Storage resized and monitored

SMTP authentication failed

Invitations and reset emails stop sending

An SMTP relay changed its auth method, or a sender domain lost DKIM, and the server quietly stops delivering invitations. New users cannot onboard.

Mail flowing, invitations landing

nginx upstream closed connection

Clients time out intermittently

The web vault, the API container or the identity service is restarting under load, often because a container was given too little memory for the vault size.

Containers sized for real use

What we do

Bitwarden help, from self-hosted to SCIM

Engage us for one broken server, or hand over the whole business deployment.

Rescue
Server not syncing, SCIM out of sync, emergency access stuck, or a self-hosted install that has drifted off a supportable version. We get it running first and plan the long path afterward.
Self-host deployment
A clean install of either the multi-container or Unified setup, with real backups, real TLS and real monitoring rather than the demo compose file.
SCIM and Directory Connector
Provisioning from Okta, Azure or Google wired correctly, with a handler that treats inbound events as idempotent rather than fighting duplicates.
Upgrade path
Walking a lagging install through the supported sequence of versions, testing clients and the database each step of the way.
Migration from a hosted plan
Moving from the hosted Bitwarden tier to self-hosted, or in the other direction, without losing shares or breaking SSO.
Managed Tech
Monthly health checks on the containers, storage and queues, plus a line to call when a client reports a sync error at the worst possible moment.

How it works

From first message to fixed

  1. Look Free estimate

    Tell us what's going wrong. We look at it, then tell you plainly what's wrong and what it will cost.

  2. Fix

    We fix the cause, not the symptom. If something is down, getting it back up comes first.

  3. Harden

    Backups, monitoring, updates and access, so the same problem can't quietly come back.

  4. Hand back or stay on

    You get a plain-language write-up. Or we stay on for hosting, upkeep and new work.

Often found together

Running Bitwarden alongside something else?

We work on the whole stack, so the problem can be wherever it actually is.

Questions

Bitwarden questions

Something else? Call+1 737 727 7327

Should we self-host Bitwarden?

Only if you have the operational appetite for it. Self-hosting gives you data residency and offline control but adds backups, TLS, SMTP, database and upgrade tasks to your plate. Teams with a mature platform group tend to do well; teams without one often end up running an aging install that nobody wants to touch.

Why do our self-hosted clients keep failing to sync?

Three common reasons: the API container is out of memory, the attachments or SQL volume is nearly full, or the database connection pool is exhausted during peak hours. We watch the containers through a sync cycle, confirm which one is the bottleneck, and resize or retune the piece that is the actual constraint.

How do we move from the hosted plan to self-hosted?

There is a supported export and import path, but it has caveats around shared organizations, attachments and sends. We script the export, stand up the target server, run a staged import, and move users over in groups rather than trying to migrate everyone in one night.

Why is SCIM creating duplicate Bitwarden users?

Almost always because the externalId the IdP is sending does not match what the server stored previously, either from a provisioning toggle or from a reconfiguration in Okta or Azure. We set a stable externalId attribute in the IdP and clean up the duplicates on the server side before re-enabling.

What happens with emergency access if the primary user leaves?

The trusted contact submits an access request, waits the configured period, and then can view or take over the vault depending on what you set. The flow only works if the trusted contact is still an active account and knows how to use the feature. We document the runbook and test it before you need it.

How much does Bitwarden help cost?

Our quote is free. Send over the server logs or the symptom, and we price a fixed engagement before touching the install.

Free estimate

Bitwarden giving you trouble?

Tell us what you're running and what's going wrong. A person reads it and comes back with an estimate.

  • No charge to find out what it will take
  • If it's urgent, say so

Replies come from a person. By sending, you authorize us to email you back. No marketing.