Scale

`sfp server scale`

Scale worker services for a tenant

`sfp server scale`

Adjust the number of worker instances for different queue priorities to handle varying workloads.

USAGE
  $ sfp server scale -t <value> [--json] [--passphrase <value>
    [--identity-file <value> --ssh-connection <value>]] [--critical-workers
    <value>] [--normal-workers <value>] [--batch-workers <value>] [-g
    <value>...] [--loglevel
    trace|debug|info|warn|error|fatal|TRACE|DEBUG|INFO|WARN|ERROR|FATAL]

FLAGS
  -t, --tenant=<value>              (required) Name of the tenant
  --critical-workers=<value>        Number of critical queue workers (1-10)
  --normal-workers=<value>          Number of normal queue workers (1-10)
  --batch-workers=<value>           Number of batch queue workers (1-10)
  
  SSH OPTIONS
  --ssh-connection=<value>          SSH connection string in the format user@host[:port]
  --identity-file=<value>           Path to SSH private key file
  --passphrase=<value>              Passphrase for the SSH private key if required
  
  OTHER OPTIONS
  --json                            Format output as json
  -g, --logsgroupsymbol=<value>...  Symbol used by CICD platform to group/collapse logs
  --loglevel=<option>               [default: info] logging level for this command invocation
                                    <options: trace|debug|info|warn|error|fatal|TRACE|DEBUG|INFO|WARN|ERROR|FATAL>

DESCRIPTION
  Scale worker services for a tenant

  Worker Types:
  - Critical Workers: Handle high-priority, time-sensitive tasks (auth, critical operations)
  - Normal Workers: Process standard operations and deployments
  - Batch Workers: Execute background jobs, scheduled tasks, and non-urgent operations

  Each worker type can be scaled independently from 1 to 10 instances.

EXAMPLES
  $ sfp server scale --tenant my-app --critical-workers 2

  $ sfp server scale --tenant my-app --normal-workers 3

  $ sfp server scale --tenant my-app --batch-workers 2

  $ sfp server scale --tenant my-app --critical-workers 2 --normal-workers 3 --batch-workers 2

  $ sfp server scale --tenant my-app --critical-workers 2 --ssh-connection [email protected] --identity-file ~/.ssh/id_rsa

Worker Types

Critical Workers

Handle high-priority, time-sensitive operations:

Authentication requests
Security operations
System-critical tasks
Real-time notifications

sfp server scale --tenant my-app --critical-workers 3

Normal Workers

Process standard operations:

Package deployments
Build operations
Standard API requests
Regular task processing

sfp server scale --tenant my-app --normal-workers 5

Batch Workers

Execute background and scheduled tasks:

Report generation
Data cleanup
Scheduled maintenance
Long-running operations

sfp server scale --tenant my-app --batch-workers 2

Scaling Strategies

Scale Up for High Load

Increase workers during peak times:

# Scale up all worker types
sfp server scale --tenant production \
  --critical-workers 5 \
  --normal-workers 8 \
  --batch-workers 4

Scale Down for Maintenance

Reduce workers for maintenance:

# Minimal configuration
sfp server scale --tenant production \
  --critical-workers 1 \
  --normal-workers 1 \
  --batch-workers 0

Gradual Scaling

Scale workers gradually to test capacity:

# Step 1: Increase normal workers
sfp server scale --tenant my-app --normal-workers 3

# Step 2: Monitor performance
sfp server status --tenant my-app

# Step 3: Adjust if needed
sfp server scale --tenant my-app --normal-workers 5

Output Formats

Standard Output

Scaling workers for tenant: my-app

Current Configuration:
  Critical Workers: 1 → 3
  Normal Workers:   2 → 5
  Batch Workers:    1 → 2

✓ Scaled critical workers to 3
✓ Scaled normal workers to 5
✓ Scaled batch workers to 2

New Configuration:
  Critical Workers: 3 (running)
  Normal Workers:   5 (running)
  Batch Workers:    2 (running)

Scaling completed successfully

JSON Output

{
  "tenant": "my-app",
  "status": "success",
  "previous": {
    "critical": 1,
    "normal": 2,
    "batch": 1
  },
  "current": {
    "critical": 3,
    "normal": 5,
    "batch": 2
  },
  "workers": {
    "critical": {
      "requested": 3,
      "running": 3,
      "status": "healthy"
    },
    "normal": {
      "requested": 5,
      "running": 5,
      "status": "healthy"
    },
    "batch": {
      "requested": 2,
      "running": 2,
      "status": "healthy"
    }
  },
  "timestamp": "2024-01-15T10:30:00Z"
}

Remote Server Scaling

Scale workers on a remote server:

sfp server scale --tenant production \
  --critical-workers 3 \
  --ssh-connection [email protected] \
  --identity-file ~/.ssh/production_key

Scaling Recommendations

Based on Load Patterns

Light Load (< 100 operations/hour):

sfp server scale --tenant my-app \
  --critical-workers 1 \
  --normal-workers 1 \
  --batch-workers 1

Medium Load (100-1000 operations/hour):

sfp server scale --tenant my-app \
  --critical-workers 2 \
  --normal-workers 3 \
  --batch-workers 2

Heavy Load (> 1000 operations/hour):

sfp server scale --tenant my-app \
  --critical-workers 4 \
  --normal-workers 6 \
  --batch-workers 3

Auto-Scaling Script

#!/bin/bash
# auto-scale.sh

TENANT="my-app"
THRESHOLD_HIGH=80
THRESHOLD_LOW=20

# Get current queue sizes
QUEUE_STATUS=$(sfp server status --tenant $TENANT --json)
NORMAL_QUEUE=$(echo $QUEUE_STATUS | jq '.queues.normal.pending')

if [ $NORMAL_QUEUE -gt $THRESHOLD_HIGH ]; then
  echo "High load detected, scaling up..."
  sfp server scale --tenant $TENANT --normal-workers 5
elif [ $NORMAL_QUEUE -lt $THRESHOLD_LOW ]; then
  echo "Low load detected, scaling down..."
  sfp server scale --tenant $TENANT --normal-workers 2
fi

Performance Monitoring

Monitor worker performance after scaling:

# Check worker status
sfp server status --tenant my-app --json | jq '.workers'

# Monitor queue sizes
sfp server logs --tenant my-app --service normal-worker --tail 50

# Check resource usage
sfp server health --detailed

Best Practices

Start Conservative: Begin with fewer workers and scale up as needed
Monitor After Scaling: Always check status after scaling operations
Consider Queue Depth: Scale based on queue backlog, not just time
Balance Worker Types: Ensure appropriate ratio between worker types
Plan for Peaks: Pre-scale before expected high-load periods

Troubleshooting

Workers Not Starting

# Check logs for errors
sfp server logs --tenant my-app --service critical-worker --tail 100

# Verify resources
sfp server health --detailed

High Memory Usage

# Scale down workers
sfp server scale --tenant my-app \
  --critical-workers 1 \
  --normal-workers 2 \
  --batch-workers 1

Note: Worker scaling changes take effect immediately but may take a few moments for all instances to be fully operational.

Warning: Scaling down workers will gracefully shutdown excess instances, but any running tasks will be completed first.

PreviousLogs NextApplication Token

Last updated 3 days ago

Was this helpful?