SQL Server Error 35206

Error number

35206

Severity

10 (informational)

State

Not documented by Microsoft (not reproduced here)

Product

Microsoft SQL Server Database Engine

Versions

SQL Server 2016 to 2025 with Always On availability groups

Category

High Availability

Last reviewed

24 October 2026

What it means

Message 35206 in the SQL Server error log means a replica stopped answering pings on an existing availability group connection for longer than the session timeout, so the connection to that replica was treated as failed.

Replicas in an availability group keep a connection open between their database mirroring endpoints and ping each other. Each replica has a session timeout (10 seconds by default). If no ping response arrives within it, the connection is declared failed and 35206 is written to the error log of the replica that noticed it, naming the other replica and its id. While the connection is down, data movement to that secondary stops and it shows as DISCONNECTED.

Severity 10 means this is an informational log entry, not an error returned to applications. The message offers two explanations: a network or firewall problem between the servers, or the other replica changing to the RESOLVING role (for example during a failover or a cluster problem). A single entry followed by 35202 a few seconds later is a short interruption; repeated entries point at an unstable network, an overloaded server or a session timeout that is set too low.

Common causes

• A network interruption or packet loss between the replicas, or a firewall or security appliance closing idle or long-lived connections on the endpoint port.
• The other replica was too busy to answer in time: very high CPU, memory pressure, a memory dump being written or a virtual machine being paused or throttled.
• The other replica changed role to RESOLVING, for example during an automatic failover, a lease timeout or loss of cluster quorum.
• A session timeout lower than the default of 10 seconds, which Microsoft warns can make a busy system miss pings and declare a false failure.

Diagnose it

Read-only These queries only read metadata and dynamic management views. They do not change data, permissions or configuration, so they are safe to run first.

SELECT ar.replica_server_name, rs.role_desc, rs.connected_state_desc,
       rs.synchronization_health_desc, rs.last_connect_error_number,
       rs.last_connect_error_description, rs.last_connect_error_timestamp
FROM sys.availability_replicas AS ar
LEFT JOIN sys.dm_hadr_availability_replica_states AS rs ON rs.replica_id = ar.replica_id;

SELECT ag.name AS ag_name, ar.replica_server_name, ar.session_timeout,
       ar.availability_mode_desc, ar.failover_mode_desc, ar.endpoint_url
FROM sys.availability_replicas AS ar
JOIN sys.availability_groups AS ag ON ag.group_id = ar.group_id;

SELECT e.name, e.state_desc, e.role_desc, t.port, e.connection_auth_desc, e.encryption_algorithm_desc
FROM sys.database_mirroring_endpoints AS e
JOIN sys.tcp_endpoints AS t ON t.endpoint_id = e.endpoint_id;

SELECT DB_NAME(drs.database_id) AS database_name, ar.replica_server_name,
       drs.synchronization_state_desc, drs.is_suspended, drs.suspend_reason_desc,
       drs.log_send_queue_size, drs.redo_queue_size
FROM sys.dm_hadr_database_replica_states AS drs
JOIN sys.availability_replicas AS ar ON ar.replica_id = drs.replica_id;

How to fix it

Each fix is labelled with what it changes. Try read-only checks and the least invasive fix first, and test changes outside production.

ALTER AVAILABILITY GROUP SalesAg
MODIFY REPLICA ON N'SQL02' WITH (SESSION_TIMEOUT = 10);

Examples

Finding the entries in the error log

Shown as documented in the message catalogue; not reproduced here because it needs an availability group. The message appears in the SQL Server error log of the replica that lost the connection, not in a query window. The replica name and id are examples.

-- Search the current error log for the message
EXEC sys.xp_readerrorlog 0, 1, N'35206';

A connection timeout has occurred on a previously established connection to availability replica 'SQL02' with id [4A0E6C38-1B7E-4E1F-9C3A-2F6D1E8B7A10].  Either a networking or a firewall issue exists or the availability replica has transitioned to the resolving role.

How to prevent it

Keep the session timeout at 10 seconds or more, give replica traffic a reliable network path, and exclude the endpoint port from idle connection timeouts on firewalls between the servers. Alert on connected_state_desc = DISCONNECTED and on growing send queues, so short interruptions are noticed and repeated ones are investigated before they cause a failover.

Common questions

Learn more on CodeWithSQL

• RDS Multi-AZ and read replicas

Official documentation

• Change the session-timeout period for an availability replica
• The database mirroring endpoint
• Lease, cluster and health check timeouts for availability groups
• sys.dm_hadr_availability_replica_states (Transact-SQL)

Last technically reviewed 24 October 2026; examples and diagnostics run on SQL Server 2025 (17.0.1145.4). Message text checked against the engine's sys.messages catalogue.