backblaze b2 / amazon s3 as remote in xoa

nraynaud

@shorian I would like to abuse your patience again, by asking you to test this branch: https://github.com/vatesfr/xen-orchestra/tree/nr-s3-fix-big-backups2

The concept is that the backup upload will happen without any sort of smart upload system or queue.

Thank you, Nico

shorian

@nraynaud Building XO now; will run a few backups overnight.

Anything in particular you'd like me to look for / test ?

nraynaud

@shorian thanks. If you could just try no concurrency as a proof that the code work, then your normal concurrency, to check if the issue is fixed that would be great.

This code is meant to be more vulnerable to losing the connection to S3, but I think it's a low risk. It's now using straight HTTP POST.

shorian

@nraynaud Bit of a hiccup I'm afraid ; errors out with "Not Implemented":

{
  "data": {
    "mode": "full",
    "reportWhen": "failure"
  },
  "id": "1615570162008",
  "jobId": "b947a31c-35f7-45d9-af88-628ec027c71e",
  "jobName": "B2",
  "message": "backup",
  "scheduleId": "8481e1b8-3d22-4502-abb7-9f0413aca0a3",
  "start": 1615570162008,
  "status": "pending",
  "infos": [
    {
      "data": {
        "vms": [
          "25a4e2a6-b116-7cee-b94b-f3553fe001f2",
          "24103ce1-e47b-fe12-4029-d643e0382f08",
          "60ac886d-8adf-9e21-4baa-b14e0fc1b2bb"
        ]
      },
      "message": "vms"
    }
  ],
  "tasks": [
    {
      "data": {
        "type": "VM",
        "id": "25a4e2a6-b116-7cee-b94b-f3553fe001f2"
      },
      "id": "1615570162050",
      "message": "backup VM",
      "start": 1615570162050,
      "status": "pending",
      "tasks": [
        {
          "id": "1615570162064",
          "message": "snapshot",
          "start": 1615570162064,
          "status": "success",
          "end": 1615570163265,
          "result": "294d65fc-ddc7-b8ad-68b8-d14e36b6838e"
        },
        {
          "data": {
            "id": "df07dd6b-753e-40f2-89be-75404bac2c1e",
            "type": "remote",
            "isFull": true
          },
          "id": "1615570163286",
          "message": "export",
          "start": 1615570163286,
          "status": "failure",
          "tasks": [
            {
              "id": "1615570164234",
              "message": "transfer",
              "start": 1615570164234,
              "status": "failure",
              "end": 1615570164257,
              "result": {
                "message": "Not implemented",
                "name": "Error",
                "stack": "Error: Not implemented\n    at S3Handler._createWriteStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/fs/src/abstract.js:428:11)\n    at S3Handler._createOutputStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/fs/src/abstract.js:412:25)\n    at S3Handler.createOutputStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/fs/src/abstract.js:129:12)\n    at RemoteAdapter.outputStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/RemoteAdapter.js:513:34)\n    at /opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/_FullBackupWriter.js:70:7\n    at FullBackupWriter.run (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/_FullBackupWriter.js:69:5)\n    at Array.<anonymous> (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/_VmBackup.js:209:9)"
              }
            }
          ],
          "end": 1615570164258,
          "result": {
            "message": "Not implemented",
            "name": "Error",
            "stack": "Error: Not implemented\n    at S3Handler._createWriteStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/fs/src/abstract.js:428:11)\n    at S3Handler._createOutputStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/fs/src/abstract.js:412:25)\n    at S3Handler.createOutputStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/fs/src/abstract.js:129:12)\n    at RemoteAdapter.outputStream (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/RemoteAdapter.js:513:34)\n    at /opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/_FullBackupWriter.js:70:7\n    at FullBackupWriter.run (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/_FullBackupWriter.js:69:5)\n    at Array.<anonymous> (/opt/xo/xo-builds/xen-orchestra-202103121606/@xen-orchestra/backups/_VmBackup.js:209:9)"
          }
        }
      ]
    }
  ]
}

I'll try over the weekend with a fresh install. Backing up to Backblaze B2 rather than S3 - likely to be the issue?

nraynaud

@shorian Ah sorry, I think S3 is completely broken at the moment.

shorian

@nraynaud Getting the "not implemented" error on all branches at the moment - is there a functioning, albeit not perfect branch that we can run with as a temporary measure? Thanks chap

yomono

Is this still broken on all branches?
It's a shame, it was working relatively fine when used carefully

olivierlambert

@julien-f and @nraynaud are working on it. That's why it's still called beta But we should have soon something on track.

shorian

New S3 function (albeit using Backblaze) seems to be working really well for small backups; I'm now seeing almost 100% success.

However larger backups are consistently failing with

"message": "Error calling AWS.S3.upload: aborted",
            "name": "Error",
            "stack": "Error: Error calling AWS.S3.upload: aborted\n    at rethrow (/opt/xo/xo-builds/xen-orchestra-202105081131/node_modules/@sullux/aws-sdk/index.js:254:24)\n    at runMicrotasks (<anonymous>)\n    at processTicksAndRejections (internal/process/task_queues.js:93:5)\n    at async S3Handler._outputStream (/opt/xo/xo-builds/xen-orchestra-202105081131/@xen-orchestra/fs/dist/s3.js:100:5)\n    at async S3Handler.outputStream (/opt/xo/xo-builds/xen-orchestra-202105081131/@xen-orchestra/fs/dist/abstract.js:250:5)\n    at async RemoteAdapter.outputStream (/opt/xo/xo-builds/xen-orchestra-202105081131/@xen-orchestra/backups/RemoteAdapter.js:509:5)\n    at async /opt/xo/xo-builds/xen-orchestra-202105081131/@xen-orchestra/backups/writers/FullBackupWriter.js:69:7"

Running with xo-server 5.79.3 , xo-web 5.81.0

I'm having the same issue across different buckets in B2 and from different servers.

Any ideas?

shorian

As an aside, using the same build I am also getting problems when running CR over a slower WAN, with the following error. Just mentioning it here in case it's related.

"message": "VDI_IO_ERROR(Device I/O errors)",
            "name": "XapiError",
            "stack": "XapiError: VDI_IO_ERROR(Device I/O errors)\n    at Function.wrap (/opt/xo/xo-builds/xen-orchestra-202105081131/packages/xen-api/dist/_XapiError.js:26:12)\n    at _default (/opt/xo/xo-builds/xen-orchestra-202105081131/packages/xen-api/dist/_getTaskResult.js:24:38)\n    at Xapi._addRecordToCache (/opt/xo/xo-builds/xen-orchestra-202105081131/packages/xen-api/dist/index.js:761:51)\n    at /opt/xo/xo-builds/xen-orchestra-202105081131/packages/xen-api/dist/index.js:789:14\n    at Array.forEach (<anonymous>)\n    at Xapi._processEvents (/opt/xo/xo-builds/xen-orchestra-202105081131/packages/xen-api/dist/index.js:774:12)\n    at Xapi._watchEvents (/opt/xo/xo-builds/xen-orchestra-202105081131/packages/xen-api/dist/index.js:931:14)\n    at runMicrotasks (<anonymous>)\n    at processTicksAndRejections (internal/process/task_queues.js:93:5)"
          }
        }
      ],
      "end": 1620864312824,
      "result": {
        "message": "all targets have failed, step: writer.transfer()",
        "name": "Error",
        "stack": "Error: all targets have failed, step: writer.transfer()\n    at VmBackup._callWriters (/opt/xo/xo-builds/xen-orchestra-202105081131/@xen-orchestra/backups/_VmBackup.js:118:13)\n    at runMicrotasks (<anonymous>)\n    at processTicksAndRejections (internal/process/task_queues.js:93:5)\n    at async VmBackup._copyDelta (/opt/xo/xo-builds/xen-orchestra-202105081131/@xen-orchestra/backups/_VmBackup.js:190:5)\n    at async VmBackup.run (/opt/xo/xo-builds/xen-orchestra-202105081131/@xen-orchestra/backups/_VmBackup.js:371:9)"

CR to a local machine is working fine, the remote one is failing every iteration.

olivierlambert

Thanks for your feedback

@julien-f is working on that

shorian