Data loss in volume

January 31, 2018 2.1k views
DigitalOcean Security CentOS Storage

Guys, I’ve been talking totally frustrated! I recently decided to migrate my clients to Digital Ocean. This week I’m gradually migrating all accounts. However, something very annoying happened and DO just does not answer me any more. They responded only once with a totally out-of-context response.

I follow these steps in migration.

1) I backup to the old server.
2) I transfer the backup directly between the servers. The speed of the network is faster, if it were to download on my machine and resending would take a long time.
3) I import the account on the new server (cpanel) and move the backup to a temporary volume.
4) I delete the old server backup because I need space to generate other backups.

After the migration I intended to download the files from the volume and destroy it. It turns out that to my surprise, during the incident “129x6k9l8b0m” my droplet had deleted files. I lost tens of GB of data. To be honest I fuck#$ beautifully, I’m trying to reduce the crap, making new backups of the imported accounts. Only the first files saved on the volume, the files saved later were lost.

But what do I do? You do not have a phone to call. They do not answer my ticket, they do not talk if they can recover the data, nor do they talk about what happened! Could do refund too!

I send the server history, showing the commands that moved the backup to the volume, and NO command was executed to delete volume data.

I’m very worried about leaving my clients staying here! Have you ever had this problem? Because I lost confidence in the company = /

If someone from DO support read this and help me out I’m really grateful.

4 Answers

Now I’m in the same situation, on Wednesday we had a volume attached to a dropplet, today this volume dissapear literally. The support responds quickly but they only check the same console that I have access as a client. I checked the command history and NO command to unmount o destroy the volume.

that’s really scary and surprising nobody from digital ocean dare reply such a post

  • Samuel, they responded via ticket. They argued that it was my fault, even sending all the logs, I showed all the commands executed on the machine. Something happened, but they say it’s your fault, that the system is isolated, etc. But unfortunately I have lost confidence in them.

Is this still a problem going into 2019 ? Im scared because I have never heard of such happening with Google Cloud or AWS lightsail. Is this something that was unique at the time or continues to happen for users? Losing their data or droplet/s.

  • I was asked by you to look into this case and respond here. I’ve done so today. It was a bit difficult to piece things together all this time later, but I’ve done my best.

    This was not the issue that I thought it was in our email conversation. Rather, this appears to have no relation to an issue that we experienced with our infrastructure. I understand the user’s frustration and I do not want to make any negative suggestions in their direction at all. I do have to say that there is/was no evidence that an issue occurred on our side in relation to this event.

    Things happen and people make mistakes, maybe the user did and maybe they didn’t. I can’t say for sure. I can only say that all systems are/were functional and that no data loss issues occurred on our side at the time of this event.

  • Hello.

    So, since it was not a significant loss, I simply moved on. I continue with my systems at Digital Ocean and I have never had any problems.

    In recent years my droplets are working normally. Just like backups, etc. I’ll even stop and see all the files in the case, but that time I did not find any command that could have deleted the data from the external storage. I’ve been working with the data center for years, I think some virtualization snapshot was due for some internal error or incident, but it’s been years, it’s no time to point the finger and blame someone.

    In summary: You can use the services normally, I use the service of backup of the own DO and whenever I needed I was able to use normally. I tested Google Cloud, AWS, Heroku, etc. At all times I have voted to stay in DO.

    You’ve never heard of this problem, because it was probably a super isolated case.

Have another answer? Share your knowledge.