Is Snowflake Down? Complete Status Check Guide + Quick Fixes
Snowflake queries timing out?
Can't connect to your data warehouse?
Snowpipe not loading data?
Before escalating to support, verify if Snowflake is actually down—or if it's a configuration, network, or credential issue on your end. Here's your complete guide to checking Snowflake status and fixing common problems fast.
Quick Check: Is Snowflake Actually Down?
Don't assume it's Snowflake. 70% of "Snowflake down" reports are actually network issues, warehouse suspension, authentication problems, or query timeout errors.
1. Check Official Sources
Snowflake Status Page:
🔗 status.snowflake.com
What to look for:
- ✅ "All Systems Operational" = Snowflake is fine
- ⚠️ "Degraded Performance" = Some services affected
- 🔴 "Major Outage" = Snowflake is down
Real-time updates:
- Compute layer issues (warehouses)
- Storage layer problems
- Metadata service status
- Authentication failures
- Regional cloud outages (AWS/Azure/GCP)
Twitter/X Search:
🔗 Search "Snowflake down" on Twitter
Why it works:
- Users report outages instantly
- See if others in your region/cloud affected
- Snowflake support team responds here
Pro tip: If 200+ tweets in the last hour mention "Snowflake down," it's probably actually down.
2. Check Service-Specific Status
Snowflake has multiple layers that can fail independently:
| Service | What It Does | Status Check |
|---|---|---|
| Compute Layer | Virtual warehouses, query execution | status.snowflake.com |
| Storage Layer | Cloud storage (S3/Azure Blob/GCS) | Check status page under "Storage" |
| Metadata Service | Account info, object definitions | Check status page under "Metadata" |
| Authentication | Login, SSO, OAuth | Check status page under "Authentication" |
| Snowpipe | Continuous data loading | Check status page under "Snowpipe" |
| Data Sharing | Cross-account sharing | Check status page under "Data Sharing" |
Your service might be down while Snowflake globally is up.
How to check which service is affected:
- Visit status.snowflake.com
- Look for specific service status by region
- Check "Incident History" for recent issues
- Subscribe to status updates (email/SMS)
- Filter by your cloud provider (AWS/Azure/GCP)
3. Check Your Cloud Region
Snowflake runs on multiple cloud providers and regions:
AWS Regions:
- US East (N. Virginia, Ohio)
- US West (Oregon, N. California)
- Canada (Central)
- Europe (Frankfurt, Ireland, London, Paris, Stockholm)
- Asia Pacific (Mumbai, Seoul, Singapore, Sydney, Tokyo)
Azure Regions:
- East US 2, West US 2
- West Europe, North Europe
- Australia East, Southeast Asia
GCP Regions:
- us-central1, us-east4
- europe-west2, europe-west4
- asia-northeast1
How to check your region:
-- Check your current region
SELECT CURRENT_REGION();
-- Check your account region
SHOW REGIONS;
Region-specific outages are common. Check if other regions are working.
4. Test Different Connection Methods
If SnowSQL works but Python connector doesn't, it's likely your driver/library.
| Method | Test Command |
|---|---|
| Web UI | Login at app.snowflake.com |
| SnowSQL | snowsql -a <account> -u <username> |
| Python | import snowflake.connector; conn = snowflake.connector.connect(...) |
| JDBC | Test with your Java application or SQL client |
| ODBC | Test with Tableau, Power BI, or Excel |
Decision tree:
Web UI works + SnowSQL fails → Client/network issue
Web UI fails + SnowSQL fails → Authentication/network issue
Python works + JDBC fails → Driver version mismatch
Nothing works → Snowflake is down (or credentials expired)
Common Snowflake Error Messages (And What They Mean)
Error 250001: "Could Not Connect to Snowflake Backend"
What it means: Can't establish connection to Snowflake.
Causes:
- Network firewall blocking Snowflake
- VPN/proxy interference
- DNS resolution failure
- Wrong account identifier
Quick fixes:
- Check if app.snowflake.com loads in browser
- Verify account identifier:
<orgname>-<account_name> - Check DNS:
nslookup <account>.snowflakecomputing.com - Disable VPN temporarily
- Allow Snowflake through firewall (port 443)
Error 390144: "Authentication Token Has Expired"
What it means: Your session token expired.
Causes:
- Session idle for >4 hours (default timeout)
- SSO token expired
- Clock skew between client and Snowflake
- Key pair rotation (for key-pair auth)
Quick fixes:
- Re-authenticate (log in again)
- Check system clock is accurate
- For key-pair auth: rotate and update keys
- Increase
CLIENT_SESSION_KEEP_ALIVEparameter:
ALTER SESSION SET CLIENT_SESSION_KEEP_ALIVE = TRUE;
Error 002003: "SQL Compilation Error: Object Does Not Exist"
What it means: Table, view, or database doesn't exist or you lack access.
Causes:
- Object name misspelled
- Wrong database/schema context
- Object dropped
- Insufficient privileges
Quick fixes:
- Check object exists:
SHOW TABLES LIKE 'table_name';
SHOW DATABASES LIKE 'db_name';
- Set correct context:
USE DATABASE my_database;
USE SCHEMA my_schema;
- Check privileges:
SHOW GRANTS ON TABLE my_table;
Error 090105: "Cannot Perform Operation. Warehouse Suspended"
What it means: Virtual warehouse is not running.
Causes:
- Warehouse auto-suspended due to inactivity
- Manually suspended by admin
- Insufficient credits/spending limit reached
- Warehouse dropped
Quick fixes:
- Resume warehouse:
ALTER WAREHOUSE my_warehouse RESUME;
- Check warehouse status:
SHOW WAREHOUSES LIKE 'my_warehouse';
- Enable auto-resume:
ALTER WAREHOUSE my_warehouse SET AUTO_RESUME = TRUE;
Error 100038: "Query Execution Timeout"
What it means: Query took too long to execute.
Causes:
- Large dataset scan
- Missing indexes/clustering
- Warehouse too small
- Resource contention
- Statement timeout parameter
Quick fixes:
- Increase statement timeout:
ALTER SESSION SET STATEMENT_TIMEOUT_IN_SECONDS = 3600; -- 1 hour
- Use larger warehouse:
ALTER WAREHOUSE my_warehouse SET WAREHOUSE_SIZE = 'LARGE';
- Optimize query (add filters, partition pruning)
- Check query profile in Web UI for bottlenecks
Error 090001: "Cannot Execute Statement: Warehouse Size Too Small"
What it means: Query needs more compute resources.
Causes:
- Complex aggregations on large datasets
- Too many concurrent queries
- Memory-intensive operations
- Joins on large tables
Quick fixes:
- Scale up warehouse:
ALTER WAREHOUSE my_warehouse SET WAREHOUSE_SIZE = 'X-LARGE';
- Use multi-cluster warehouse for concurrency:
ALTER WAREHOUSE my_warehouse SET
MIN_CLUSTER_COUNT = 1
MAX_CLUSTER_COUNT = 5
SCALING_POLICY = 'STANDARD';
- Optimize query to reduce memory footprint
Error 000606: "IP Address Not Allowed"
What it means: Your IP is blocked by network policy.
Causes:
- Network policy restricting IP ranges
- VPN IP not whitelisted
- Cloud function IP changed
- Dynamic IP changed
Quick fixes:
- Check current network policies:
SHOW NETWORK POLICIES;
DESC NETWORK POLICY my_policy;
- Temporarily disable (ACCOUNTADMIN only):
ALTER ACCOUNT UNSET NETWORK_POLICY;
- Add your IP to allowed list:
ALTER NETWORK POLICY my_policy SET
ALLOWED_IP_LIST = ('192.168.1.100/32', '10.0.0.0/8');
Quick Fixes: Snowflake Not Working?
Fix #1: Verify Warehouse Status
Virtual warehouses consume credits. Suspended warehouse = no queries run.
Check warehouse status:
SHOW WAREHOUSES;
Look for:
state: SUSPENDED, STARTED, RESIZINGsize: XSMALL, SMALL, MEDIUM, LARGE, etc.auto_resume: TRUE/FALSE
Resume warehouse:
ALTER WAREHOUSE my_warehouse RESUME;
-- Or resume when first query runs
ALTER WAREHOUSE my_warehouse SET AUTO_RESUME = TRUE;
Why warehouses suspend:
- Auto-suspend after X minutes of inactivity (default: 10 min)
- Manually suspended to save credits
- Credit limit reached
Best practice: Enable auto-resume for production warehouses.
Fix #2: Check Authentication & Credentials
Authentication failures are common after password changes, MFA setup, or SSO changes.
Test authentication:
SnowSQL:
snowsql -a <account> -u <username> -p <password>
Python:
import snowflake.connector
conn = snowflake.connector.connect(
user='<username>',
password='<password>',
account='<account>',
warehouse='<warehouse>',
database='<database>',
schema='<schema>'
)
# Test connection
cur = conn.cursor()
cur.execute("SELECT CURRENT_VERSION()")
print(cur.fetchone())
Common auth issues:
1. Wrong account identifier:
- Format:
<orgname>-<account_name>(new format) - Legacy:
<account_locator>.<region>.<cloud>(still supported)
Check your account:
SELECT CURRENT_ACCOUNT();
SELECT CURRENT_ORGANIZATION_NAME();
2. Password expired:
- Snowflake enforces password rotation policies
- Reset password via Web UI
3. SSO issues:
- Test SSO login via Web UI first
- Check with identity provider (Okta, Azure AD, etc.)
- Verify SSO configuration not changed
4. Key-pair authentication:
# Using key-pair auth (more secure for apps)
import snowflake.connector
conn = snowflake.connector.connect(
user='<username>',
account='<account>',
private_key=private_key_bytes,
warehouse='<warehouse>'
)
Fix #3: Test Network Connectivity
Snowflake needs port 443 (HTTPS) access to multiple endpoints.
Required endpoints:
# Test DNS resolution
nslookg <account>.snowflakecomputing.com
# Test HTTPS connectivity
curl -v https://<account>.snowflakecomputing.com
# Test authentication endpoint
curl -v https://<account>.snowflakecomputing.com/session/v1/login-request
Firewall requirements:
| Endpoint | Port | Purpose |
|---|---|---|
<account>.snowflakecomputing.com |
443 | Main account endpoint |
<account>.<region>.snowflakecomputing.com |
443 | Regional endpoint |
snowflakecomputing.com |
443 | Global services |
*.cloudfront.net (AWS) |
443 | Static content |
Corporate networks:
- Snowflake doesn't support proxy servers for drivers
- Use VPN or direct internet connection
- Contact IT to whitelist Snowflake endpoints
VPN issues:
- Some VPNs block Snowflake traffic
- Try disconnecting VPN temporarily
- Configure split tunneling to exclude Snowflake
Fix #4: Update Snowflake Connectors/Drivers
Outdated connectors = authentication failures and missing features.
Check versions:
Python:
pip show snowflake-connector-python
Current version (as of Feb 2026): 3.7.x
Update:
pip install --upgrade snowflake-connector-python
JDBC:
- Download latest from Snowflake JDBC Drivers
- Current version: 3.15.x
- Replace JAR file in your project
ODBC:
- Download from Snowflake ODBC Drivers
- Version: 3.2.x
- Reinstall driver
SnowSQL:
# Check version
snowsql -v
# Update
brew upgrade snowsql # Mac
choco upgrade snowsql # Windows
Why update matters:
- Security patches
- Bug fixes
- Support for new authentication methods (OAuth, key-pair)
- Performance improvements
Fix #5: Check Role & Privileges
"Object does not exist" often means "you can't see it."
Check current role:
SELECT CURRENT_ROLE();
Switch to role with proper access:
USE ROLE ACCOUNTADMIN; -- Full admin (use carefully!)
USE ROLE SYSADMIN; -- Warehouse/database admin
USE ROLE PUBLIC; -- Minimal access
Check granted roles:
SHOW GRANTS TO USER current_user();
Grant necessary privileges:
-- Grant warehouse usage
GRANT USAGE ON WAREHOUSE my_warehouse TO ROLE my_role;
-- Grant database access
GRANT USAGE ON DATABASE my_db TO ROLE my_role;
GRANT USAGE ON SCHEMA my_db.my_schema TO ROLE my_role;
-- Grant table access
GRANT SELECT ON TABLE my_db.my_schema.my_table TO ROLE my_role;
Common privilege issues:
- Default role changed
- Privileges revoked by admin
- Using wrong role for task
- Database/schema not accessible to role
Fix #6: Optimize Query Performance
Slow queries ≠ Snowflake down. Often it's inefficient SQL or undersized warehouse.
Check query performance:
1. Use Query Profile in Web UI:
- Login → History → Click query
- See execution timeline
- Identify bottlenecks (network, I/O, CPU)
2. Check warehouse load:
-- See currently running queries
SELECT * FROM TABLE(INFORMATION_SCHEMA.QUERY_HISTORY())
WHERE EXECUTION_STATUS = 'RUNNING'
ORDER BY START_TIME DESC;
3. Common performance issues:
a) Full table scan:
-- Bad: scans entire table
SELECT * FROM large_table WHERE date = '2026-02-01';
-- Good: use clustering key or partitions
SELECT * FROM large_table
WHERE date = '2026-02-01'
AND cluster_key_column = 'value';
b) Warehouse too small:
-- Scale up temporarily
ALTER WAREHOUSE my_warehouse SET WAREHOUSE_SIZE = 'LARGE';
-- Run query
-- Scale back down
ALTER WAREHOUSE my_warehouse SET WAREHOUSE_SIZE = 'SMALL';
c) No result cache:
-- Enable result caching (default: ON)
ALTER SESSION SET USE_CACHED_RESULT = TRUE;
d) Clustering issues:
-- Check clustering health
SELECT SYSTEM$CLUSTERING_INFORMATION('my_table', '(date_column)');
-- Manually cluster if needed
ALTER TABLE my_table CLUSTER BY (date_column);
Fix #7: Snowpipe Issues
Snowpipe not loading data? Check status and logs.
Check Snowpipe status:
-- See all pipes
SHOW PIPES;
-- Check specific pipe
SHOW PIPES LIKE 'my_pipe';
-- Get pipe status
SELECT SYSTEM$PIPE_STATUS('my_pipe');
Check load history:
-- See recent loads
SELECT * FROM TABLE(INFORMATION_SCHEMA.COPY_HISTORY(
TABLE_NAME => 'my_table',
START_TIME => DATEADD(hours, -24, CURRENT_TIMESTAMP())
));
Common Snowpipe issues:
1. Stage files not detected:
- Check S3 event notifications configured
- Verify IAM permissions for Snowflake
- Check file format matches pipe definition
2. Pipe paused:
-- Resume pipe
ALTER PIPE my_pipe REFRESH;
-- If stalled, recreate notification
ALTER PIPE my_pipe SET PIPE_EXECUTION_PAUSED = FALSE;
3. Authentication expired:
- Regenerate AWS IAM keys
- Update storage integration
- Check Azure SAS token expiry
4. File format errors:
-- Check error logs
SELECT * FROM TABLE(VALIDATE(my_pipe, JOB_ID => '_last'));
Fix #8: Data Sharing Problems
Can't access shared data? Check provider and consumer settings.
Check shares (Provider):
-- See all shares
SHOW SHARES;
-- Check share details
DESC SHARE my_share;
-- Check which accounts have access
SHOW GRANTS TO SHARE my_share;
Check shares (Consumer):
-- See available shares
SHOW SHARES;
-- Create database from share
CREATE DATABASE shared_db FROM SHARE provider_account.my_share;
-- Grant access to roles
GRANT IMPORTED PRIVILEGES ON DATABASE shared_db TO ROLE my_role;
Common sharing issues:
- Share revoked by provider
- Account identifier wrong
- Privileges not granted to roles
- Cross-region sharing limitations (same cloud provider required)
Snowflake Warehouse Not Starting?
Issue: Warehouse Stuck in "RESIZING" State
Troubleshoot:
1. Check for long-running queries:
-- Kill blocking queries
SELECT SYSTEM$CANCEL_QUERY('<query_id>');
2. Force warehouse to suspend and resume:
ALTER WAREHOUSE my_warehouse SUSPEND;
ALTER WAREHOUSE my_warehouse RESUME;
3. Check for resource limits:
- Verify credit quota not exceeded
- Check warehouse size limits in account
4. Contact Snowflake support if stuck >30 minutes
Issue: "Insufficient Resources" Error
Causes:
- Account credit limit reached
- Too many concurrent warehouses
- Warehouse size not available in region
- Resource policy restrictions
Fixes:
1. Check account credits:
-- Requires ACCOUNTADMIN
SHOW PARAMETERS LIKE 'RESOURCE%' IN ACCOUNT;
2. Reduce concurrent warehouses:
- Suspend unused warehouses
- Share warehouses between teams (use roles)
3. Use smaller warehouse:
ALTER WAREHOUSE my_warehouse SET WAREHOUSE_SIZE = 'XSMALL';
Snowflake Connection Timeout?
Issue: Connection Drops During Long Queries
Causes:
- Client timeout too short
- Network instability
- Firewall dropping long connections
Fixes:
1. Increase client timeout:
Python:
conn = snowflake.connector.connect(
...,
network_timeout=300, # 5 minutes
login_timeout=60
)
JDBC:
connectionTimeout=300000 # 5 minutes
loginTimeout=60000
2. Enable keep-alive:
ALTER SESSION SET CLIENT_SESSION_KEEP_ALIVE = TRUE;
ALTER SESSION SET CLIENT_SESSION_KEEP_ALIVE_HEARTBEAT_FREQUENCY = 3600; -- 1 hour
3. Use asynchronous queries for long operations:
# Submit query asynchronously
cur = conn.cursor()
cur.execute_async("SELECT * FROM huge_table")
# Check status
query_id = cur.sfqid
status = conn.get_query_status(query_id)
# Fetch results when ready
if conn.is_still_running(conn.get_query_status(query_id)):
print("Still running...")
else:
results = cur.fetchall()
Snowflake Connector-Specific Issues
Python Connector Issues
Error: "Failed to Get OCSP Response"
Causes:
- Firewall blocking OCSP requests
- OCSP server down
- Certificate validation failure
Fixes:
1. Bypass OCSP check (NOT recommended for production):
import os
os.environ['SNOWFLAKE_OCSP_FAIL_OPEN'] = 'true'
conn = snowflake.connector.connect(...)
2. Update connector:
pip install --upgrade snowflake-connector-python
3. Configure proxy for OCSP:
conn = snowflake.connector.connect(
...,
ocsp_response_cache_filename='/tmp/ocsp_cache'
)
JDBC Connector Issues
Error: "SSL Peer Unverified"
Causes:
- Certificate chain incomplete
- Java keystore missing Snowflake CA
- JDBC driver version mismatch
Fixes:
1. Update JDBC driver to latest version
2. Ensure JVM has proper certificates:
# Verify Java version
java -version
# Update Java if < Java 8u161
3. Add Snowflake CA to Java keystore:
keytool -import -trustcacerts -file snowflake.crt \
-alias snowflake -keystore $JAVA_HOME/jre/lib/security/cacerts
ODBC Connector Issues
Error: "Data Source Name Not Found"
Causes:
- DSN not configured
- ODBC driver not installed
- Wrong DSN name
Fixes:
1. Check ODBC driver installed:
Windows:
- ODBC Data Source Administrator (64-bit)
- System DSN tab → Check for Snowflake
Mac/Linux:
odbcinst -q -d
2. Create DSN configuration:
Windows: Use ODBC Administrator GUI
Mac/Linux: Edit ~/.odbc.ini:
[Snowflake]
Driver = /usr/lib/snowflake/odbc/lib/libSnowflake.dylib
Server = <account>.snowflakecomputing.com
UID = <username>
PWD = <password>
Database = <database>
Warehouse = <warehouse>
Schema = <schema>
3. Test connection:
isql -v Snowflake
Snowpark Issues
Issue: Snowpark Session Fails to Create
Causes:
- Outdated snowflake-snowpark-python package
- Authentication issues
- Missing dependencies
Fixes:
1. Update Snowpark:
pip install --upgrade snowflake-snowpark-python
2. Create session with error handling:
from snowflake.snowpark import Session
connection_parameters = {
"account": "<account>",
"user": "<username>",
"password": "<password>",
"warehouse": "<warehouse>",
"database": "<database>",
"schema": "<schema>",
"role": "<role>"
}
try:
session = Session.builder.configs(connection_parameters).create()
print(f"Snowpark session created: {session.sql('SELECT CURRENT_VERSION()').collect()}")
except Exception as e:
print(f"Error creating session: {e}")
3. Check Python version (requires Python 3.8+):
python --version
Issue: Snowpark UDF Fails to Run
Causes:
- Missing package dependencies
- Python version mismatch
- Warehouse not running
- Insufficient privileges
Fixes:
1. Specify packages explicitly:
from snowflake.snowpark.functions import udf
@udf(packages=['pandas', 'numpy'], is_permanent=True, name="my_udf", replace=True)
def my_function(x):
return x * 2
2. Check UDF was created:
SHOW USER FUNCTIONS LIKE 'my_udf';
3. Grant usage privileges:
GRANT USAGE ON FUNCTION my_udf(NUMBER) TO ROLE my_role;
Regional Outages: Is It Just My Cloud?
Snowflake runs on multiple cloud providers. Outages are often cloud-specific.
Check cloud provider status:
AWS: 🔗 status.aws.amazon.com
Azure: 🔗 status.azure.com
GCP: 🔗 status.cloud.google.com
How to check for cloud-specific issues:
1. Check DownDetector:
🔗 downdetector.com/status/snowflake
Shows:
- Real-time outage reports
- Heatmap of affected regions
- Spike in reports = likely real outage
2. Test from different region:
-- Replicate to different region for testing
CREATE DATABASE my_db_backup_region AS CLONE my_db;
3. Check if specific to cloud provider:
- Ask teams on different clouds (AWS vs Azure vs GCP)
- Check Snowflake status page filter by cloud
When Snowflake Actually Goes Down
What Happens
Recent major outages:
- November 2023: 4-hour AWS us-east-1 outage (S3 dependency)
- June 2023: 2-hour authentication service disruption
- March 2023: Metadata service degradation (query compilation slow)
Typical causes:
- Cloud provider outages (AWS S3, Azure Storage, GCP)
- Metadata service failures
- Authentication service issues
- Network infrastructure problems
- Software deployment bugs
How Snowflake Responds
Communication channels:
- status.snowflake.com - Primary source
- @SnowflakeDB on Twitter/X
- Email alerts (if subscribed to status page)
- In-app notifications
Timeline:
- 0-15 min: Users report issues on Twitter/Slack
- 15-30 min: Snowflake acknowledges on status page
- 30-90 min: Updates posted every 20-30 min
- Resolution: Usually 1-4 hours for major outages
What to Do During Outages
1. Check if failover possible:
- Use database replicas in different regions
- Switch to secondary Snowflake account
- Use cached results if available
2. Communicate with stakeholders:
- Share status page link
- Provide ETA from Snowflake
- Plan alternate workflows
3. Monitor status page:
- status.snowflake.com
- Subscribe to updates (email/SMS)
4. Document impact:
- Note which queries/pipelines affected
- Prepare incident report
- Review SLA credits if applicable
Snowflake Down Checklist
Follow these steps in order:
Step 1: Verify it's actually down
- Check Snowflake Status
- Check API Status Check
- Search Twitter: "Snowflake down"
- Test Web UI login at app.snowflake.com
Step 2: Quick fixes (if Snowflake is up)
- Check warehouse status (
SHOW WAREHOUSES) - Resume warehouse if suspended
- Verify credentials (test SnowSQL connection)
- Check network connectivity (ping account endpoint)
- Verify role and privileges
Step 3: Connector troubleshooting
- Update connectors to latest version
- Test different connection methods (Web UI, SnowSQL, Python)
- Check for authentication errors
- Verify account identifier format
- Test with key-pair auth if using SSO
Step 4: Performance troubleshooting
- Check query profile in Web UI
- Look for long-running queries blocking warehouse
- Optimize SQL (add filters, use clustering)
- Scale up warehouse if needed
- Check for result cache misses
Step 5: Escalation
- Check cloud provider status (AWS/Azure/GCP)
- Review recent account changes
- Contact Snowflake support: community.snowflake.com
- Engage your Snowflake account team
Prevent Future Issues
1. Set Up Monitoring & Alerts
Don't wait for users to report issues.
Monitor key metrics:
-- Warehouse credit usage
SELECT
WAREHOUSE_NAME,
SUM(CREDITS_USED) AS total_credits,
COUNT(*) AS query_count
FROM SNOWFLAKE.ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY
WHERE START_TIME >= DATEADD(day, -7, CURRENT_TIMESTAMP())
GROUP BY WAREHOUSE_NAME
ORDER BY total_credits DESC;
-- Failed queries
SELECT
ERROR_CODE,
ERROR_MESSAGE,
COUNT(*) AS error_count
FROM SNOWFLAKE.ACCOUNT_USAGE.QUERY_HISTORY
WHERE START_TIME >= DATEADD(day, -1, CURRENT_TIMESTAMP())
AND ERROR_CODE IS NOT NULL
GROUP BY ERROR_CODE, ERROR_MESSAGE
ORDER BY error_count DESC;
Set up alerts:
- Use API Status Check for uptime monitoring
- Subscribe to Snowflake Status
- Monitor credit consumption thresholds
- Alert on query failure spikes
2. Implement Auto-Resume & Auto-Suspend
Prevent "warehouse suspended" errors and save credits.
ALTER WAREHOUSE my_warehouse SET
AUTO_SUSPEND = 600 -- Suspend after 10 min idle
AUTO_RESUME = TRUE -- Auto-start on query
INITIALLY_SUSPENDED = TRUE; -- Start suspended
Best practices:
- Production warehouses: AUTO_RESUME = TRUE
- Dev/test warehouses: Lower AUTO_SUSPEND (60-300 seconds)
- Batch processing: Manually control suspend/resume
3. Use Resource Monitors
Prevent unexpected credit consumption.
-- Create resource monitor (ACCOUNTADMIN required)
CREATE RESOURCE MONITOR my_monitor WITH
CREDIT_QUOTA = 1000 -- 1000 credits per month
FREQUENCY = MONTHLY
START_TIMESTAMP = '2026-02-01 00:00'
TRIGGERS
ON 75 PERCENT DO NOTIFY -- Alert at 75%
ON 100 PERCENT DO SUSPEND -- Suspend at 100%
ON 110 PERCENT DO SUSPEND_IMMEDIATE; -- Kill queries at 110%
-- Apply to warehouses
ALTER WAREHOUSE my_warehouse SET RESOURCE_MONITOR = my_monitor;
4. Optimize Query Performance
Prevent timeouts and reduce costs.
Best practices:
1. Use clustering keys for large tables:
ALTER TABLE large_table CLUSTER BY (date_column, category);
2. Partition pruning:
-- Bad: scans all partitions
SELECT * FROM sales WHERE amount > 1000;
-- Good: uses partition column
SELECT * FROM sales
WHERE date >= '2026-02-01' AND amount > 1000;
3. Materialized views for repeated aggregations:
CREATE MATERIALIZED VIEW sales_summary AS
SELECT date, region, SUM(amount) as total_sales
FROM sales
GROUP BY date, region;
4. Result caching:
-- Already enabled by default, but verify
ALTER SESSION SET USE_CACHED_RESULT = TRUE;
5. Multi-Region Failover
For critical workloads, replicate across regions.
-- Replicate database to another region
ALTER DATABASE my_db ENABLE REPLICATION TO ACCOUNTS org_name.account2;
-- Refresh replica
ALTER DATABASE my_db_replica REFRESH;
-- Promote replica to primary (failover)
ALTER DATABASE my_db_replica PRIMARY;
Considerations:
- Additional costs for replication
- Replication lag (minutes to hours)
- Same cloud provider required
- Manual failover process
Key Takeaways
Before assuming Snowflake is down:
- ✅ Check Snowflake Status
- ✅ Test Web UI at app.snowflake.com
- ✅ Verify warehouse is running (
SHOW WAREHOUSES) - ✅ Search Twitter for "Snowflake down"
Common fixes:
- Resume suspended warehouse (fixes 40% of issues)
- Update connectors/drivers to latest version
- Check authentication (password, SSO, key-pair)
- Verify network connectivity and firewall rules
- Optimize slow queries (clustering, filtering, larger warehouse)
Connector issues:
- Python: Update snowflake-connector-python
- JDBC: Download latest driver JAR
- ODBC: Reinstall driver, check DSN configuration
- SnowSQL: Update via package manager
Performance issues:
- Use Query Profile to identify bottlenecks
- Scale warehouse up temporarily for heavy queries
- Implement clustering keys for large tables
- Enable result caching (on by default)
If Snowflake is actually down:
- Monitor status.snowflake.com
- Check cloud provider status (AWS/Azure/GCP)
- Usually resolved within 1-4 hours
- Review SLA for credit compensation
Prevent future issues:
- Enable AUTO_RESUME on production warehouses
- Set up resource monitors for credit limits
- Monitor query failures and warehouse usage
- Implement multi-region replication for critical data
Remember: Most "Snowflake down" issues are suspended warehouses, authentication problems, or network/firewall issues. Check these first before assuming a platform outage.
Need real-time Snowflake status monitoring? Track Snowflake uptime with API Status Check - Get instant alerts when Snowflake goes down.
Related Resources
- Is Snowflake Down Right Now? — Live status check
- Snowflake Outage History — Past incidents and timeline
- Snowflake vs Databricks Uptime — Which platform is more reliable?
- API Outage Response Plan — How to handle downtime like a pro
Monitor Your APIs
Check the real-time status of 100+ popular APIs used by developers.
View API Status →