Bug Report

I'm currently on the most recent RDM 2026.2.4.4. This issue has existed even prior to this release also. When I quit RDM and select "Yes" for the "Are you sure you want to close Remote Desktop Manager" prompt, I will get a MacOS notification noise (not actual notification, just the noise like when you click tab in a terminal and it's telling you it's done the autocomplete). During this time, it seems to be pasting a bunch of gibberish text in a massive dump of uknown data. I'm not sure exactly what's in it, jsut that it's doing it. Often, I'll notice later that my MacOS hard drive is renamed because the random gibbersish pasted after the exit somehow managed to select my hard drive (presumabily hit Enter, and rename it with text). I have no idea what would cause this or how the applciation still has permissions to type after it's exited, but it's somewhow escaping the application sandbox and pasting text outside of the application itself.

Hi, Running RDM for macOS 2026.2.4.4 on macOS 15.7.8, I encountered an issue where when configuring a user-specific override to select a credential object from my user vault, trying to select the session entry in the tree after editing it, results in the UI resetting to the first available (in alphabetic order) vault in the list and related dashboard. If I remove the user-specific override, the behaviour returns to normal. The issue only seems to be with session entries, not with folders (there it works as expected). I can also trigger the issue when the browser plug-in (using Edge for macOS with integration configured to RDM) tries to access the same session type to fill in the login info. In the background, I see RDM switching to the first vault too. When attempting to set the user-specific override in Windows, the macOS version again exhibits the problem so it doesn't seem to be related to which client version is making the change. If debug information is needed, please let me know, and I'll provide what I can. Best regards, Jelle Hillen

I have an automated process that creates a JSON for my hosts and i import into RDM and it works great. This for a new build/standup for a customer. It's a new customer every week and then i just delete it and re-create for the next customer and then re-import into a new database. I'm trying to login to Linux systems via SSH. I use the broadcast feature to then send multiple setup commands to each host to build them all in parallell. However, it seems that the Post-Login page of the SSH session is not working. First, i'm trying to send 2x CTRL-C to break the login process and return me to a command prompt. It sends the text {CTRL}C but not the actual control character. I might be doing it wrong. What is the string to add in Post-Login to be able to send a CTRL-C? (or actually send 2 CTRL-C with a 1 second delay after initial SSH is established, and then another 1 second delay until the 2nd CTRL-C.) Second, when i try to edit what i have, i cannot add new commands. The + button is is greyed out. I can edit the existing commands added via JSON, and i can edit the delay. But i cant seem to add new commands. Here is the JSON i create to import. It will create the SSH session and link to a linked credential. { "ConnectionType": "SSHShell", "OpenEmbedded": true, "CredentialConnectionID": "xxxxxxxxxxxxxxxxxxxxx", "Group": "xxx\\xxx\\xxx\\xxx", "TabGroupMode": "Inherited", "ColorMode": "Inherited", "Name": "xxxxxxxxxx", "Terminal": { "UseServerWindowTitle": false, "Font": "Menlo", "DisconnectAction": "KeepOpen", "PrivateKeyPromptForPassPhrase": false, "EnableLogging": true, "ColorMode": 1, "LogOverwriteMode": "Append", "AlwaysAcceptFingerprintDefaultBoolean": true, "Username": "xxxx", "Host": "xxxxx", "MaxScrollbackLines": 9999, "ColorANSIBlue": "#0086FD", "ColorDefaultForeground": "#F0F8FF", "FontMac": "SF Mono", "AlwaysAcceptFingerprint": true, "FontSize": 13, "ColorDefaultBackground": "#270603", "FontMode": "Override", "LogMode": "AllPrintableOutput", "AfterConnectTerminalMacros": [ { "Macro": "{CTRL}C" }, { "Macro": "{CTRL}C" } ] }, "VPN": { "Mode": 5 } }, (Sorry, when i pasted the JSON it removed all the indenting for some reason) Here is the GUI and the + button is greyed out. [image]

Hi everyone, I'm currently facing an issue with Devolutions RDM Free on macOS (version 2026.2.4.4) . I have a host configured with a username, password, and OTP. I have also correctly filled in the OTP Prompt String . The problem is that on macOS, despite having the OTP Prompt String configured, the application still manually prompts me for the One-Time Password. This exact same configuration works perfectly fine on Windows for the exact same host (the OTP is injected automatically). It only seems to fail on the macOS version. Has anyone else experienced this issue, or is there an additional setting I need to check on macOS? Thanks in advance for any help!

Hello when MAC switch to dark mode, the ribbon become unusable, label of ribbon group are white. Thanks [image]

In the latest macOS RDM release (2026.2.4.4), I have observed an issue with tab reordering when the total width of open tabs exceeds the available width of the tab bar. When a tab is located beyond the visible portion of the tab bar and is selected, it is correctly displayed on the far right side of the visible tab area. This behavior is expected and appears to be functioning as designed. However, once the tab bar enters this overflow state, tabs can no longer be reliably rearranged. As long as the total tab width exceeds the available pixel width of the tab bar, drag-and-drop reordering either does not work at all or behaves unexpectedly. In some cases, attempting to move a tab causes it to automatically reposition itself as the second tab from the left. I'm not sure why this specific location is chosen, but it is consistently what I've observed. After the tab is moved to that position, it cannot be moved back to its original location and effectively becomes stuck. I previously reported Tab Not Displayed for Current Tab if Single Row Width is Enabled , which was subsequently fixed. This new issue may be related to the changes made for that fix, although I cannot confirm that. It is also possible that this tab reordering problem existed prior to this release and simply went unnoticed because of the previous tab visibility issue. Expected Behavior Tabs should remain fully draggable and reorderable regardless of whether the tab bar has exceeded its available width. Selecting a tab that is outside the visible area should not affect the ability to reposition tabs. Actual Behavior Once the tab bar overflows the available width, tab reordering becomes unreliable or stops working entirely. Attempting to move a tab may unexpectedly relocate it to the second position from the left. Tabs that have been relocated in this manner cannot be moved back to their intended position. [image]

We do not know what to do anymore... RDM is constantly failing with error "unable to retrieve 1password credentials". we file a new ticket, they tipically wait a month to fix it, and one or two weeks later it stops working again... are we the only ones using 1password with RDM? is there any reccomandation? Many thanks in advance to who can help us.

The feature to double-click to close tabs is no longer working. I have it enabled in the settings, but the tabs only close when right-click > Close is selected or the X of th tab is pressed. I have the Double-Click action set to Close, but it's no longer working in the current version (2026.2.1.5). [image]

When a tab is open it should be actively shown on the tab bar. When it is outside the single row width of available pixels, it should still display the current tab, even if it woudl otherwise overflow into the unseen tabs from the single row. With the latest RDM redesign (2026.2.1.5), if I have a tab selected that is outside the curent single row width, it is not displayed. In previous releases, if it was outside the single-row width, it would render the current viewed tab to the far-right of the visible and avaialble pixel area of the single-width bar. With the latest release, this is no longer happening. [image] If I enable "Show tabs with multiple rows," the tab is visible (as expected). [image]

Hello after updating RDP latest build on a old macbook with Ventura 13.7.8, it freeze during loadin connection from devolution server. Rollback to previous version work. Try the cfg rename but unsussesfully. Any ideas? Thanks

Environment: RDM for macOS 2026.2.3.1 macOS 26.5.2 (25F84), Apple Silicon Embedded RDP session Spotlight “Results from Clipboard” disabled Summary: With an embedded RDP session active, taking a local screenshot using Control+Command+Shift+4 causes RDM to replace the valid PNG on the macOS clipboard with promised Windows image formats. When a local application requests the advertised TIFF, RDM fails in PasteboardCliprdr.ToTiff() and never fulfills the clipboard promise. The receiving application then hangs through repeated 60-second timeouts. This affects Slack, Chrome/Jira, Paintbrush and other Mac applications. It reproduces with Maccy or ClipBook and sometimes with no clipboard manager running. Steps to reproduce: Open an embedded RDP session in RDM. On the local Mac, press Control+Command+Shift+4 and select an area. Paste into a local application such as Slack. The receiving application hangs or beachballs. If allowed to time out, Slack may show image.png with a “not the right format” error. Expected: The local screenshot should paste normally. RDM should not replace it with an image representation it cannot render. Diagnostic findings: The unified macOS pasteboard log captures the complete failure: screencapture writes a 168,832-byte public.png. RDM successfully reads all 168,832 bytes. About 330 ms later, RDM becomes the new clipboard owner. RDM replaces the PNG with promised formats: com.devolutions.rdp.cf-metafilepict com.devolutions.rdp.cf-dib com.devolutions.rdp.cf-dibv5 public.tiffSlack requests public.tiff. pboard asks RDM to fulfill the promise. RDM throws: Could not initialize an instance of the type 'AppKit.NSBitmapImageRep': the native 'initWithData:' method returned nil. at AppKit.NSBitmapImageRep..ctor(NSData data) at Devolutions.Rdp.Native.PasteboardCliprdr.ToTiff(...) at Devolutions.Rdp.Native.PasteboardCliprdr.Provider_OnRenderClipboard(...) at Devolutions.Rdp.Native.PasteboardItemProvider.ProvideDataForType(...) RDM does not fulfill the promise. Slack times out after 60 seconds and retries several times, remaining blocked for approximately four minutes. A Slack process sample confirms its main thread is blocked in: NSImage initWithPasteboard NSPasteboard _dataForType CFPasteboardCopyData mach_msg2_trapA simultaneous pboard sample shows pboard idle while waiting for the promised-data provider. The RDM sample shows its application and FreeRDP threads otherwise responsive. The failure occurs specifically in RDM’s macOS clipboard image conversion. Scope: Multiple unrelated receiving applications are affected. Maccy and ClipBook can increase the frequency but are not required. Disabling Spotlight clipboard results does not help. Disabling RDP clipboard redirection or disconnecting RDM avoids the problem. Saving the screenshot to a file and dragging it into the application also works. Related fixes: RDM Mac 2026.1.12.2, released May 20, claimed improved macOS clipboard-history support to avoid RDP session hangs. A related forum report was marked Resolved/Implemented, although Devolutions later said the original changes addressed symptoms rather than the root cause: RDM macOS clipboard freeze report RDM Mac release notes RDM 2026.2.2.1 also claimed to fix hangs when copying RDP text with a third-party clipboard manager: 2026.2.2.1 announcement This case is different because it involves a local macOS PNG, affects local receiving applications and can occur without a clipboard manager. The logs point directly to Devolutions.Rdp.Native.PasteboardCliprdr.ToTiff().

Hello ! It feels like Apple broke something with their latest developer beta of Golden Gate. RDM is crashing with specific popup windows. 1) Opening the about screen. Instant crash. 2) Opening any connection that needs credentials. As soon as the credential window starts to load - crash. 3) Opening the properties of a connection without saved credentials. Selecting the RDP section and then pressing TAB to go to the username filed will crash RDM.

Environment Remote Desktop Manager 2026.2.1.2 (64-bit), macOS (Apple Silicon) Embedded RDP engine: …/Remote Desktop Manager.app/Contents/MonoBundle/libDevolutionsRdp.dylib Connection type: RDP (Microsoft Remote Desktop) through an RD Gateway (WebSocket-based, open-source rdpgw) Gateway server: stage-rdpgw.oneclick.services:443 (valid Let's Encrypt cert, reachable, /remoteDesktopGateway/ returns HTTP 401 as expected) Connection Host : a per-session hostname (e.g. u3wsva2djj-stage.oneclick.services) that is only resolvable through the gateway , not via local/public DNS (by design — the gateway resolves it internally) Problem Every connection attempt fails within ~3 seconds with: ERROR Could not establish a secure connection to the host (0x00020008) (occasionally Connection failed (0x00020006)). Nothing ever reaches the gateway server (verified server-side: no handshake/auth is logged). Root cause (from RDM debug+verbose log) RemoteDesktopManager.debug shows, in the same second as the error: DEBUG Network monitoring initialized for host: u3wsva2djj-stage.oneclick.services DEBUG Network transition detected for host: u3wsva2djj-stage.oneclick.services - Pausing connections DEBUG Certificate verification skipped: Network transition in progress → ERROR Could not establish a secure connection to the host (0x00020008) The network monitor then loops endlessly , emitting dozens of identical lines per second and never settling: DEBUG Network restored for host: … - Starting stabilization period DEBUG Network transition detected for host: … - Pausing connections It appears the network monitor performs a reachability check tied to the connection's Host , and because that Host is only resolvable via the gateway (not local DNS), the check fails permanently → it perpetually reports a "network transition" and pauses the connection, so the actual connect never proceeds (TLS/cert verification is explicitly skipped). Proof this is RDM-specific (not network/gateway/target) The native Microsoft RDP client (Windows App / mstsc) connects successfully to the same gateway + same Host from the same Mac and network . A FreeRDP 3.x -based client also connects successfully to the same gateway. Gateway endpoint reachable from the Mac (TCP/TLS OK via openssl/nc); macOS Application Firewall is disabled ; no Little Snitch / network-filter extensions. This connection worked previously (January 2026) on an older RDM version (which used the separate FreeRDPSandboxed.app engine and did not have this network-monitoring behavior). On 2026.2.x no SessionLogs/*.log is written for these attempts — the engine aborts before a session starts. Already tried (no effect) Clean reinstall of RDM 2026.2.1.2 Recreating the connection entry from scratch Gateway Transport = Auto and = HTTP Stopping Tailscale/VPN (the monitor kept flapping for 90s+ afterwards) No user-facing toggle for this in RemoteDesktopManager.cfg Questions / request How can the RDP network-monitoring / "network transition" pausing behavior be disabled? This looks like a false-positive when the RDP Host is only resolvable through the RD Gateway (common with gateway-brokered/per-session hostnames). Can the monitor be fixed to not block such connections (e.g. check the gateway reachability instead of the inner Host)?

Version: RDM macOS 2026.1.12.2, Devolutions.ProtocolsSharp 2026.4.17.1 Server: OpenSSH 8.9p1 Ubuntu 22.04, xterm-256color Behavior: Session establishes successfully, runs for varying duration (seconds to minutes), then unexpectedly terminates with the above exception. Server logs show clean Received disconnect: Logout (client-side disconnect). Log excerpt attached. Stack: DoShellLoopAsync: Arg_IndexOutOfRangeException at Devolutions.Protocols.Core.Terminal.ExecuteCommand(ReadOnlySpan`1 data) at Devolutions.Protocols.Core.Terminal.ParseXtermData(ReadOnlySpan`1 data) at Devolutions.Protocols.Core.SshClient.DoShellLoopAsync(Channel channel, Upper upper)

HI, I am experiencing frequent issues with RDP sessions in Remote Desktop Manager, where the application becomes unresponsive when attempting to copy content from an active RDP session. This issue occurs consistently in approximately two out of every three attempts. When it happens, the application remains unresponsive for a couple of minutes before recovering and functioning normally again. I am currently running the latest version (2026.1.10.3) on an M5 MacBook Pro with the latest version of macOS. Please note that this is a new setup on this MacBook, and the .rdm configuration file was imported from my previous MacBook.

With the latest 2026.2.1.5 RDM for MacOS, I cannot open any of my saved connections when I double click on them. When I double click on them, I receive a message stating "The action needs an open CyberArk dashboard with an account selected." I have never configured CyberArk on my RDM settings and I checked in all the obvious places and CyberArk isn't enabled, both within the RDM settings and also the individual connection settings. [image] I am still able to open connections if I left-click to select the session and press the "Enter" key and also through the search shortcut of the application (Shift + CMD + \). This is why I think it's a bug and not a newly implemented feature change since I can still open it through the other two methods.

I am using RDM on an 2017 iMac running MacOS Ventura 13.7.8. Before the upgrade RDM ran fine. After upgrading to 2026.2.1.5, RDM got stuck with the loading connections part. See attachment. I removed it from my computer, reinstalled it, no matter what I am doing it always ends up hanging at the same step. Does anybody have the same problem? If so, has anybody been able to overcome it? Thank you any comments in advance [image]

How solve this bug when loading devolutions cloud in rdm for macOS in latest version 2026.2.1.2 Unable to convert entry to connection. ID: fdc22095. Name: Devolutions account validation. Exception: Devolutions.Cryptography.DevolutionsCryptoException: NativeError : InvalidMac ===Original Message=== Exception_WasThrown, Devolutions.Cryptography.DevolutionsCryptoException ===Original StackTrace=== at Devolutions.Cryptography.Utils.HandleError(Int64 errorCode) at Devolutions.RemoteDesktopManager.Managers.DevolutionsHubDataSourceManager.ConvertEntryToCommonConnection(Guid dataSourceID, Guid repositoryID, Entry entry) at Devolutions.RemoteDesktopManager.Managers.DevolutionsHubDataSourceManager.ConvertEntryToConnection(Guid dataSourceID, Guid repositoryID, Entry entry, Dictionary`2 entries) at Devolutions.RemoteDesktopManager.Managers.DevolutionsHubDataSourceManager.GetConnections(ConnectionEngine engine, Guid dataSourceID, Guid repositoryID, Guid dataSourceVerison, Int64 repositoryVersion)` ——— Unable to convert entry to connection. ID: 3a0a3c29. Name: Authenticator App. Exception: Devolutions.Cryptography.DevolutionsCryptoException: NativeError : InvalidMac ===Original Message=== Exception_WasThrown, Devolutions.Cryptography.DevolutionsCryptoException ===Original StackTrace=== at Devolutions.Cryptography.Utils.HandleError(Int64 errorCode) at Devolutions.Cryptography.Managed.Decrypt(Byte[] data, Byte[] key, Byte[] aad, ILegacyDecryptor legacyDecryptor) at Devolutions.Cryptography.Managed.DecryptWithKeyAsUtf8String(String b64data, Byte[] key, Byte[] aad, ILegacyDecryptor legacyDecryptor) at Devolutions.RemoteDesktopManager.Managers.DevolutionsHubDataSourceManager.ConvertEntryToCommonConnection(Guid dataSourceID, Guid repositoryID, Entry entry) at Devolutions.RemoteDesktopManager.Managers.DevolutionsHubDataSourceManager.ConvertEntryToConnection(Guid dataSourceID, Guid repositoryID, Entry entry, Dictionary`2 entries) at Devolutions.RemoteDesktopManager.Managers.DevolutionsHubDataSourceManager.GetConnections(ConnectionEngine engine, Guid dataSourceID, Guid repositoryID, Guid dataSourceVerison, Int64 repositoryVersion)`

Hi, Not really a bug report but with the new UI changes in RDM 2026.2 on macOS, the close buttons for tabs are now on the right side of the tab instead of the left side, which is rather counter-intuitive for Mac users. Just saying. [image] Best regards, Jelle Hillen

Hey, I've been trying to establish multiple embedded RDP sessions with the latest Remote Desktop Manager for MacOS (2026.1.8.6), but none of them worked. I tried to establish "plain" RDP sessions (direct connection to the target machine), but it didn't work. I also tried to establish CyberArk's SIA-RDP connections (with RD Gateway), but it didn't work either. The message I keep getting is "unable to connect to host - CONNECT_FAILED (0x00020006)". I also tried to change the hostname to the public IP, but it didn't work as well. The thing is - I tried all of those sessions with an external RDP client (Windows App triggered via Remote Desktop Manager), and all of them worked. I asked a team member to try on their machine and they got the same issue. The way I configured my RDP sessions was by importing RDP files, which, again, worked well on Windows App. My setup: MacOS Tahoe (26.3.1) Latest Remote Desktop Manager version (2026.1.8.6) Palo Alto Networks VPN

Since latest update (2026.1.12.4) terminal tool struggling to start and if it starts it is starting with the error 22 as yo can see in the screenshot. [image] And it started but the first line has percent sign and only after first enter key it fixes it's place. [image] Thanks

Hello Devolutions Support Team, I would like to report a bug in Remote Desktop Manager macOS version 2026.1.12.4 affecting the 1Password integration when DUO MFA is enabled. **Environment:** - RDM macOS: 2026.1.12.4 - 1Password for Mac: 8.10.21 - 1Password CLI: installed and up to date - 1Password account type: Business (baselnetgroup.1password.com) - MFA method: DUO - RDM 1Password config: My Account Settings, Host: Default, Region: Global, Login method: Default **Issue:** When attempting to retrieve credentials from 1Password via RDM, the following error is displayed: "Unable to retrieve 1Password credentials" The Application Log reveals the underlying cause: "PasswordManagerAccess.Common.InternalErrorException: Duo: Duo HTML: the main form is not found" RDM appears to be attempting to handle the DUO MFA flow through its own embedded WebView rather than delegating authentication to the 1Password CLI / 1Password App integration. The "Connect with 1Password CLI" option is enabled in the 1Password App under Settings → Developer, but RDM does not appear to honour this and still tries to process DUO authentication internally — which fails because the current DUO HTML no longer contains the form element RDM expects. **Steps to reproduce:** 1. Configure 1Password in RDM via My Account Settings (Region: Global, Login method: Default) 2. Enable "Connect with 1Password CLI" in 1Password App → Settings → Developer 3. Attempt to use a credential entry linked to 1Password in RDM 4. Observe the error **Workaround:** Using a 1Password Service Account token instead of username/password/MFA resolves the issue, but this is not acceptable as a permanent solution for our organisation due to the reduced security level (no MFA on the service account token). **Expected behaviour:** RDM should delegate authentication to the 1Password CLI / App integration and not attempt to handle DUO MFA via an internal WebView. Please advise on a fix or patch timeline. Best regards, Lukas Schweizer

I use multiple RDP connections over SSH tunnels with VPN default pauser time set to 0 seconds . This worked flawlessly up to 2025.2.12.6 . Even if the SSH tunnel needed ~1 second to fully establish, RDP would automatically connect once ready. Since newer versions, RDP connections only succeed if I manually set at least 1 second pause , even for tunnels that establish instantly. With pause=0, the SSH tunnel connects, but RDP never starts. Additionally, when using SSH tunnels with TOTP + private key only (without static password): if I enter the OTP in the One Time Password setting, it is ignored and I am still prompted for an OTP code. Authentication only succeeds if I enter the OTP into the password field . This issue does not occur when starting the SSH tunnel alone — in that case, the TOTP field works correctly and is automatically used.

If I open / edit and close multiple files on a server that have the same name (eg docker-compose.yml) then whenever one of those files is "saved" in a local editor (VS Code and Mac TextEdit), RDM then automatically uploads the single file being "saved" to the destinations of any files that were opened & closed in the session with the same file name effectively overwriting them with the incorrect content. I am on version 26.1.10.3, this behavior is new. Previously i could open and edit multiple files at the same time. I believe the cause is because the local cache directory where the files are temporarily stored does not take into account of the path of the file on the remote server. IE it just stores docker-compose.yml at the base cache directory. Hope that makes sense? [image]

[image] I'am running the lateset RDM for Mac (2026.1.9.5) and startying from previous relese (2026.1.8.7) stop working change directory on SFTP session Now RDM shown only user home dir and not allow change it (event for root user) Direct sftp to remote server works fine Now in UI in the path if i can type any existing directory on remote server after Enter nothing happen And in local side SFTP broser now showing / instead home dir