자동 롤백
Compute Engine에서 VM을 목표 수만큼 만들 수 없으면 요청이 자동으로 롤백됨
아니요
사용 설정(선택사항)
API 비율 제한
요청이 API 비율 제한에 미치는 영향 정도
VM당 요청 하나
각 일괄 작업에 대한 요청 하나
제한사항
다음 목록은 VM을 일괄 생성할 때 적용되는 제한사항을 설명합니다.
고객 제공 암호화 키.고객 제공 암호화 키(CSEK)로 보호되는 디스크가 있는 VM을 만들 수 없습니다.
디스크는 Google 기본값 또는 고객 관리 암호화 키를 사용해야 합니다.
스냅샷을 사용해 부팅 디스크 만들기. 스냅샷을 사용하여 VM의 부팅 디스크를 만들 수는 없습니다. 대신 VM을 일괄 생성할 때 커스텀 이미지를 사용하세요.
전역 DNS. 전역 DNS를 사용하는 경우 정규화된 도메인 이름(FQDN)에 영역이 포함되어 있지 않으므로 이름이 충돌할 수 있습니다. 이를 방지하려면 영역 DNS를 사용합니다. 영역 DNS로 전환할 수 없으면 여러 리전에서 이름이 같은 패턴을 사용하지 마세요. 자세한 내용은 내부 DNS 이름 유형을 참조하세요.
[[["이해하기 쉬움","easyToUnderstand","thumb-up"],["문제가 해결됨","solvedMyProblem","thumb-up"],["기타","otherUp","thumb-up"]],[["이해하기 어려움","hardToUnderstand","thumb-down"],["잘못된 정보 또는 샘플 코드","incorrectInformationOrSampleCode","thumb-down"],["필요한 정보/샘플이 없음","missingTheInformationSamplesINeed","thumb-down"],["번역 문제","translationIssue","thumb-down"],["기타","otherDown","thumb-down"]],["최종 업데이트: 2025-05-29(UTC)"],[[["\u003cp\u003eBulk VM creation allows for the creation of numerous identical and independent virtual machines with a single API request, effectively minimizing API rate limit consumption.\u003c/p\u003e\n"],["\u003cp\u003eWhen creating VMs in bulk, Compute Engine can automatically distribute the VMs across multiple zones within a region, optimizing for resource availability and honoring any existing reservations.\u003c/p\u003e\n"],["\u003cp\u003eUnlike batching API requests, bulk VM creation offers automatic zone selection, upfront validation of requests, optional automatic VM name generation, and the possibility of automatic rollback if VM creation fails.\u003c/p\u003e\n"],["\u003cp\u003eThere are limitations to bulk VM creation, such as not supporting customer-supplied encryption keys, boot disks from snapshots, machine images, sole-tenant nodes, and requiring most VM properties to be identical.\u003c/p\u003e\n"],["\u003cp\u003eUsing bulk VM creation does not incur additional charges, and billing only starts upon successful VM creation; additionally, audit logs are provided for both the bulk operation and individual VMs.\u003c/p\u003e\n"]]],[],null,["Linux Windows\n\n*** ** * ** ***\n\nBulk creation of virtual machine (VM) instances lets you create a large number\nof VMs that are identical and independent from each other. When you create VMs\nin bulk, you can do the following:\n\n- **Minimize consumption of API rate limits.** Requests for creating VMs in bulk\n consume [API rate limits](/compute/docs/api-rate-limits) as if you were\n creating a single VM.\n\n- **Distribute VM creation across multiple zones in a region.**\n For regional requests, Compute Engine determines the zone to create\n the VMs in based on the zones that have available hardware, taking into\n consideration the available capacity in each zone, as well as any of your\n [reservations](/compute/docs/instances/choose-reservation-type).\n\nCompute Engine doesn't manage VMs that you create in bulk. To create a\ngroup of managed VMs, use [managed instance groups](/compute/docs/instance-groups).\n\nCreating VMs in bulk versus batching API requests\n\nCreating VMs in bulk is different from [batching API\nrequests](/compute/docs/api/how-tos/batch) with the [`instances.insert`\nmethod](/compute/docs/reference/rest/v1/instances/insert). The following table\ncompares these two ways of creating multiple VMs:\n\n| Functionality | Batching API requests | Creating VMs in bulk |\n|-------------------------------------------------------------------------------------------------------------------|-------------------------|--------------------------------------------|\n| **Zone selection** Automatically selected based on considerations such as resource availability and quota | Manual | Automatic when using the regional endpoint |\n| **Upfront validation** Request fails immediately if it is not feasible | No | With capacity and quota |\n| **VM name generation** Automatically generated based on a specified name pattern | Manual | Optionally generated automatically |\n| **Automatic rollback** Request automatically rolled back if Compute Engine cannot create the target number of VMs | No | Optionally enabled |\n| **API rate limit** How requests affect the API rate limit | One request for each VM | One request for each bulk operation |\n\nLimitations\n\nThe following list describes the limitations that apply when creating VMs in\nbulk:\n\n- **Customer supplied encryption keys.** You can't create VMs that have disks\n that are protected by [customer-supplied encryption keys (CSEK)](/compute/docs/disks/customer-supplied-encryption).\n The disks must use Google-default or customer-managed encryption keys.\n\n- **Boot disks from snapshots.** You can't use a snapshot to create the boot disks\n for the VMs. Instead, use [a custom image](/compute/docs/images/create-custom#create_image)\n when you create VMs in bulk.\n\n- **Global DNS.** If you are using global DNS, there might be name conflicts\n because the fully qualified domain name (FQDN) doesn't contain the zone. To\n avoid this, use zonal DNS. If you cannot switch to zonal DNS, avoid using the\n same name pattern across different regions. For more information, see [Types\n of internal DNS names](/compute/docs/internal-dns#instance-fully-qualified-domain-names).\n\n- **Machine images.** You cannot create VMs from [machine\n images](/compute/docs/machine-images).\n\n- **Number of VMs.** You can create up to 5,000 VMs with each call. Also, you\n can create a maximum of 8,000 instances simultaneously at any time. This\n simultaneous operation limit is shared between\n [`instances.insert`](/compute/docs/reference/rest/v1/instances/insert) and\n [`instances.bulkInsert`](/compute/docs/reference/rest/v1/instances/bulkInsert).\n\n- **Sole-tenant nodes.** You cannot create VMs that use\n [sole-tenant node affinity labels](/compute/docs/nodes/sole-tenant-nodes#node_affinity_and_anti-affinity).\n\n- **VM properties.** All [VM properties](/compute/docs/instances/update-instance-properties),\n except their names and custom hostnames, must be identical. You cannot use\n VM properties that are mutually exclusive between VMs, which include but\n are not limited to static external IP addresses and static internal IP addresses.\n\nAudit logs\n\nCompute Engine logs information about the bulk VM creation to the\n[Admin Activity audit log](/logging/docs/audit#admin-activity) when you make the\nAPI call and when the API call completes.\n\nCompute Engine also creates separate audit logs for each VM. You can\nfind the audit log for a single VM by matching the value of\n`protoPayload.resourceName` with the VM name generated by the naming pattern\nthat you specify with the request.\n\nPricing\n\nThere are no additional charges for using bulk VM creation. As with creating\nsingle VMs, billing begins when you create the VMs.\n\nYou are not billed for any VMs that Compute Engine fails to create.\n\nWhat's next\n\n- [Create VMs in bulk](/compute/docs/instances/multiple/create-in-bulk)"]]
