Skip to content

servicegroup.md

Latest commit

 

History

History
150 lines (105 loc) · 3.33 KB

servicegroup.md

File metadata and controls

150 lines (105 loc) · 3.33 KB

servicegroup - ServiceGroup Member Quorum Monitoring

Note: All examples assume environment variables are set:

export NETSCALER_HOST=192.168.1.10
export NETSCALER_USER=nsroot
export NETSCALER_PASS=nsroot

See Environment Variables for details.

Monitor ServiceGroup member availability and quorum (percentage of active members).

Basic Usage

Check servicegroup quorum

check_netscaler -C servicegroup -n web_sg -w 50 -c 25

Output (OK):

OK: servicegroup web_sg: 8/10 members UP (80%) | web_sg.member_quorum=80;50;25

Output (WARNING):

WARNING: servicegroup web_sg: 4/10 members UP (40%) | web_sg.member_quorum=40;50;25

Output (CRITICAL):

CRITICAL: servicegroup web_sg: 2/10 members UP (20%) | web_sg.member_quorum=20;50;25

Threshold Logic

Thresholds are minimum percentages of active members:

  • WARNING: Alert if quorum falls BELOW warning %
  • CRITICAL: Alert if quorum falls BELOW critical %

Example with -w 50 -c 25:

  • OK: ≥50% members UP
  • WARNING: 25-49% members UP
  • CRITICAL: <25% members UP

Advanced Usage

Custom separator for perfdata

check_netscaler -C servicegroup -n web_sg -w 50 -c 25 --separator "_"

Output:

OK: servicegroup web_sg: 8/10 members UP (80%) | web_sg_member_quorum=80;50;25

Check without thresholds

check_netscaler -C servicegroup -n web_sg

Returns OK with member count, no threshold checking.

Member States

The check counts members as UP if their state is:

  • UP - Member is operational
  • ENABLED - Member is enabled and healthy

Members are DOWN if:

  • DOWN - Member is down
  • OUT OF SERVICE - Member is disabled
  • GOING OUT OF SERVICE - Member is being drained

Performance Data

| servicegroupname.member_quorum=percentage;warn;crit;0;100

Example:

| web_sg.member_quorum=80;50;25;0;100

Common Use Cases

1. Production services (strict quorum)

check_netscaler -C servicegroup -n prod_web_sg -w 75 -c 50

Requires at least 75% healthy, critical below 50%.

2. Development services (relaxed quorum)

check_netscaler -C servicegroup -n dev_web_sg -w 25 -c 10

More tolerant of member failures.

3. Database cluster (high availability)

check_netscaler -C servicegroup -n db_cluster_sg -w 66 -c 33

2 out of 3 nodes minimum (66%), critical if only 1 node (33%).

4. Information only (no thresholds)

check_netscaler -C servicegroup -n monitoring_sg

Quorum Calculation

quorum % = (UP members / total members) × 100

Examples:

  • 8/10 members UP = 80% quorum
  • 5/10 members UP = 50% quorum
  • 1/10 members UP = 10% quorum

Exit Codes

  • 0 (OK) - Quorum above or equal to warning threshold
  • 1 (WARNING) - Quorum below warning but above/equal critical threshold
  • 2 (CRITICAL) - Quorum below critical threshold
  • 3 (UNKNOWN) - Cannot retrieve servicegroup data

Tips

  • Set warning threshold based on acceptable degraded performance
  • Set critical threshold based on minimum viable service level
  • For N+1 redundancy: warn at 100%, crit at 50%
  • For N+2 redundancy: warn at 66%, crit at 33%
  • Use with state command for detailed member status
  • ServiceGroup name is case-sensitive
  • Members are bound to services, not servers directly