IBM Cloud Support integration

Know when and how to engage IBM Cloud Support for performance issues.

Self-service resources

Before opening a ticket, check the following:

When to contact IBM Support

Contact support if the following apply:

Performance issues persist after following this guide.
Replication lag continues despite optimization.
Disk latency remains high without workload spikes.
Suspected infrastructure-level issues.
Unexpected behavior after scaling.
Deployment health issues.
Backup or restore problems.

Before contacting support

Gather the following information:

Deployment details
- Deployment ID (CRN)
- Region and availability zones
- Current plan and resources
- MongoDB version
Issue details
- Time window of the issue (with timezone)
- Symptoms observed
- Impact on applications
- Recent changes (code, configuration, scaling)
Performance data
- Monitoring screenshots from Sysdig
- Query examples causing issues
- Output from diagnostic commands
- Activity Tracker events during the issue window

MongoDB diagnostics

# Collect diagnostic data
mongo "your-connection-string" --eval "
  printjson(db.serverStatus());
  printjson(db.currentOp());
  printjson(rs.status());
" > mongodb-diagnostics.json

Opening a support ticket

Using the IBM Cloud console

Navigate to Support in the top menu.
Click Create a case.
Select Databases for MongoDB.
Choose severity level.
Provide detailed description.
Attach diagnostic files.

Using the IBM Cloud CLI

# Create support case
ibmcloud support case-create \
  --subject "MongoDB Performance Issue" \
  --description "Detailed description of issue" \
  --severity 2 \
  --offering databases-for-mongodb

Severity levels

Severity levels
Severity	Description	Response time
1 (Critical)	Production down, data loss	1 hour
2 (High)	Significant performance degradation	2 hours
3 (Medium)	Moderate impact, workaround available	4 hours
4 (Low)	General questions, feature requests	8 hours

Escalation procedures

If the issue is not resolved within the expected timeframe:

Update the support case with urgency.
Contact your IBM account team.
For critical issues, request management escalation.

Support best practices

Provide complete information upfront.
Respond promptly to support requests.
Test suggested solutions in non-production first.
Document resolution for future reference.

Quick reference: diagnostic commands

Essential MongoDB commands for performance troubleshooting.

Diagnostic commands
Command	Purpose	Key metrics	Normal values
`db.serverStatus()`	Overall server statistics	CPU, memory, connections	Varies by workload
`db.serverStatus().connections`	Connection statistics	current, available	< 80% of available
`db.serverStatus().opcounters`	Operation counters	insert, query, update, delete	Baseline dependent
`db.serverStatus().locks`	Lock statistics	Global lock time	< 10% of total time
`db.serverStatus().wiredTiger.cache`	Cache statistics	Cache hit ratio	95%
`db.currentOp()`	Current operations	Active queries, locks	Few long-running ops
`db.currentOp({ waitingForLock: true })`	Operations waiting for locks	Lock contention	Should be empty
`rs.status()`	Replica set status	Replication lag, member health	Lag < 1 second
`rs.printSecondaryReplicationInfo()`	Replication lag details	Lag per secondary	Lag < 1 second
`db.collection.stats()`	Collection statistics	Size, index size, document count	Monitor growth
`db.collection.find().explain("executionStats")`	Query execution plan	Execution time, docs examined	Use indexes
`db.system.profile.find()`	Slow query log	Slow operations	Review regularly
`sh.status()`	Sharding status (if applicable)	Chunk distribution	Even distribution
`db.adminCommand({ top: 1 })`	Collection usage statistics	Hot collections	Identify optimization targets
`db.printReplicationInfo()`	Oplog information	Oplog size, time range	Sufficient for recovery

Quick diagnostic workflow

// 1. Check overall health
db.serverStatus().ok  // Should return 1

// 2. Check connections
var conn = db.serverStatus().connections;
print("Connections: " + conn.current + "/" + conn.available);

// 3. Check replication (if replica set)
rs.status().ok  // Should return 1

// 4. Check for slow operations
db.currentOp({ "secs_running": { $gt: 5 } })

// 5. Check cache efficiency
var cache = db.serverStatus().wiredTiger.cache;
var hitRatio = 1 - (cache["pages read into cache"] /
  (cache["pages read into cache"] + cache["pages requested from the cache"]));
print("Cache hit ratio: " + (hitRatio * 100).toFixed(2) + "%");

// 6. Check for lock contention
db.currentOp({ waitingForLock: true })