Read a backup

Backups are plain zip archives, so you can read them with standard tools and without Miru. This guide downloads the latest archive, checks it, and queries it.

Download the latest archive

Archive names end in the time they were taken, so the newest archive is the last one by name.

BUCKET_NAME="<bucket-name>"
PREFIX="miru-backups" # or your custom prefix

BUCKET_NAME="<bucket-name>"
PREFIX="miru-backups" # or your custom prefix

$BucketName = "<bucket-name>"
$Prefix = "miru-backups" # or your custom prefix

AWS S3
GCS

KEY="$(aws s3api list-objects-v2 \
  --bucket "${BUCKET_NAME}" \
  --prefix "${PREFIX}/miru-backup-" \
  --query 'Contents[].Key' \
  --output text | tr '\t' '\n' | sort | tail -n 1)"
aws s3 cp "s3://${BUCKET_NAME}/${KEY}" backup.zip

KEY="$(aws s3api list-objects-v2 \
  --bucket "${BUCKET_NAME}" \
  --prefix "${PREFIX}/miru-backup-" \
  --query 'Contents[].Key' \
  --output text | tr '\t' '\n' | sort | tail -n 1)"
aws s3 cp "s3://${BUCKET_NAME}/${KEY}" backup.zip

$Key = aws s3api list-objects-v2 `
  --bucket $BucketName `
  --prefix "$Prefix/miru-backup-" `
  --query 'Contents[].Key' `
  --output text |
  ForEach-Object { $_ -split "`t" } | Sort-Object | Select-Object -Last 1
aws s3 cp "s3://$BucketName/$Key" backup.zip

URL="$(gcloud storage ls "gs://${BUCKET_NAME}/${PREFIX}/miru-backup-*.zip" | sort | tail -n 1)"
gcloud storage cp "${URL}" backup.zip

URL="$(gcloud storage ls "gs://${BUCKET_NAME}/${PREFIX}/miru-backup-*.zip" | sort | tail -n 1)"
gcloud storage cp "${URL}" backup.zip

$Url = gcloud storage ls "gs://$BucketName/$Prefix/miru-backup-*.zip" |
  Sort-Object | Select-Object -Last 1
gcloud storage cp $Url backup.zip

Above are example commands for downloading the latest archive from AWS S3 and GCS. Make sure to sign in to your cloud account before running the commands.

Unzip the archive

unzip -q backup.zip -d backup
cd backup

unzip -q backup.zip -d backup
cd backup

Expand-Archive backup.zip -DestinationPath backup
Set-Location backup

Query resources

Each line of a resource file is one resource, so jq on Linux and macOS, or PowerShell on Windows, reads the files directly. List every device’s ID and name:

jq -r '[.id, .name] | @tsv' resources/devices.jsonl

jq -r '[.id, .name] | @tsv' resources/devices.jsonl

Get-Content resources/devices.jsonl -Encoding UTF8 |
  ForEach-Object { $_ | ConvertFrom-Json } |
  ForEach-Object { "$($_.id)`t$($_.name)" }

Show one deployment, with the config instances it contains:

jq 'select(.id == "dpl_123")' resources/deployments.jsonl

jq 'select(.id == "dpl_123")' resources/deployments.jsonl

Get-Content resources/deployments.jsonl -Encoding UTF8 |
  ForEach-Object { $_ | ConvertFrom-Json } |
  Where-Object id -eq "dpl_123" |
  ConvertTo-Json -Depth 10

Read with the Python SDK

Every resource in an archive is a Platform API resource, so the Python Platform SDK models can load them. Use an SDK version built for the archive’s platform_api_version. Python reads the archive in place, without unzipping it:

import json
import zipfile

from miru_platform_sdk.types import Device

with zipfile.ZipFile("backup.zip") as archive:
    manifest = json.loads(archive.read("manifest.json"))
    print(manifest["platform_api_version"])

    with archive.open("resources/devices.jsonl") as lines:
        devices = [Device.model_validate_json(line) for line in lines]

for device in devices:
    print(device.id, device.name)

Welcome

Fleet

Config Management

Data Uploads

Developers

Administration

Read a backup

Download the latest archive

Unzip the archive

Query resources

Read with the Python SDK

​Download the latest archive

​Unzip the archive

​Query resources

​Read with the Python SDK

Download the latest archive

Unzip the archive

Query resources

Read with the Python SDK