| 1 | # you can disable an alarm notification by setting the 'to' line to: silent |
| 2 | |
| 3 | # --- Availability --- |
| 4 | |
| 5 | # Storage account availability (0-100%). Low is bad. |
| 6 | # Azure SLA guarantees >=99.9% for RA-GRS/RA-GZRS, >=99.9% for most tiers. |
| 7 | # AMBA: Availability < 99% is Sev1 (critical) |
| 8 | |
| 9 | template: am_storage_accounts_availability |
| 10 | on: azure_monitor.storage_accounts.availability |
| 11 | class: Availability |
| 12 | type: Storage |
| 13 | component: Azure Storage |
| 14 | lookup: average -5m unaligned of average |
| 15 | units: percentage |
| 16 | every: 1m |
| 17 | warn: $this < (($status >= $WARNING) ? (99.9) : (99)) |
| 18 | crit: $this < (($status == $CRITICAL) ? (99) : (90)) |
| 19 | delay: down 5m multiplier 1.5 max 1h |
| 20 | summary: Storage availability on ${label:resource_name} |
| 21 | info: Overall availability of Azure Storage Account ${label:resource_name} \ |
| 22 | in ${label:resource_group} (${label:region}). \ |
| 23 | Azure SLA guarantees 99.9% or higher depending on redundancy tier |
| 24 | to: sysadmin |
| 25 | |
| 26 | # --- Latency --- |
| 27 | |
| 28 | # End-to-end latency measures the full round-trip including network. |
| 29 | # AMBA: SuccessE2ELatency average > 1000ms is Sev2 |
| 30 | |
| 31 | template: am_storage_accounts_e2e_latency |
| 32 | on: azure_monitor.storage_accounts.e2e_latency |
| 33 | class: Latency |
| 34 | type: Storage |
| 35 | component: Azure Storage |
| 36 | lookup: average -5m unaligned of average |
| 37 | units: milliseconds |
| 38 | every: 1m |
| 39 | warn: $this > (($status >= $WARNING) ? (500) : (1000)) |
| 40 | crit: $this > (($status == $CRITICAL) ? (1000) : (2000)) |
| 41 | delay: down 5m multiplier 1.5 max 1h |
| 42 | summary: Storage E2E latency on ${label:resource_name} |
| 43 | info: Average end-to-end latency of successful requests to Azure Storage Account \ |
| 44 | ${label:resource_name} in ${label:resource_group} (${label:region}). \ |
| 45 | Includes network round-trip time |
| 46 | to: sysadmin |
| 47 | |
| 48 | # Peak end-to-end latency — catches extreme spikes even when average looks normal. |
| 49 | |
| 50 | template: am_storage_accounts_e2e_latency_peak |
| 51 | on: azure_monitor.storage_accounts.e2e_latency |
| 52 | class: Latency |
| 53 | type: Storage |
| 54 | component: Azure Storage |
| 55 | lookup: max -5m unaligned of maximum |
| 56 | units: milliseconds |
| 57 | every: 1m |
| 58 | warn: $this > (($status >= $WARNING) ? (2000) : (5000)) |
| 59 | crit: $this > (($status == $CRITICAL) ? (5000) : (10000)) |
| 60 | delay: down 5m multiplier 1.5 max 1h |
| 61 | summary: Storage E2E latency peak on ${label:resource_name} |
| 62 | info: Peak end-to-end latency of successful requests to Azure Storage Account \ |
| 63 | ${label:resource_name} in ${label:resource_group} (${label:region}). \ |
| 64 | Spikes may indicate throttling or network congestion |
| 65 | to: sysadmin |
| 66 | |
| 67 | # Server-side latency excludes network — pure Azure processing time. |
| 68 | # AMBA: SuccessServerLatency average > 1000ms is Sev2 |
| 69 | |
| 70 | template: am_storage_accounts_server_latency |
| 71 | on: azure_monitor.storage_accounts.server_latency |
| 72 | class: Latency |
| 73 | type: Storage |
| 74 | component: Azure Storage |
| 75 | lookup: average -5m unaligned of average |
| 76 | units: milliseconds |
| 77 | every: 1m |
| 78 | warn: $this > (($status >= $WARNING) ? (300) : (500)) |
| 79 | crit: $this > (($status == $CRITICAL) ? (500) : (1000)) |
| 80 | delay: down 5m multiplier 1.5 max 1h |
| 81 | summary: Storage server latency on ${label:resource_name} |
| 82 | info: Average server-side latency of successful requests to Azure Storage Account \ |
| 83 | ${label:resource_name} in ${label:resource_group} (${label:region}). \ |
| 84 | Excludes network time — high values indicate Azure-side processing delays |
| 85 | to: sysadmin |
| 86 | |
| 87 | # Peak server-side latency — catches extreme server-side spikes. |
| 88 | |
| 89 | template: am_storage_accounts_server_latency_peak |
| 90 | on: azure_monitor.storage_accounts.server_latency |
| 91 | class: Latency |
| 92 | type: Storage |
| 93 | component: Azure Storage |
| 94 | lookup: max -5m unaligned of maximum |
| 95 | units: milliseconds |
| 96 | every: 1m |
| 97 | warn: $this > (($status >= $WARNING) ? (1000) : (2000)) |
| 98 | crit: $this > (($status == $CRITICAL) ? (2000) : (5000)) |
| 99 | delay: down 5m multiplier 1.5 max 1h |
| 100 | summary: Storage server latency peak on ${label:resource_name} |
| 101 | info: Peak server-side latency of successful requests to Azure Storage Account \ |
| 102 | ${label:resource_name} in ${label:resource_group} (${label:region}). \ |
| 103 | Spikes may indicate throttling at the storage partition level |
| 104 | to: sysadmin |