Issue OneDrive Backup - Transport error - Unable to Create Backups

[httPete]

Server operating system version

Ubuntu 18.04.6 LTS

Plesk version and microupdate number

Plesk Obsidian: Version 18.0.52 Update #3

Hello all,

unfortunately I have problems with my vServer and the OneDrive backup extension in Plesk. Unfortunately, the daily backups that start at 1:30 at night are no longer saved to my OneDrive. I always get the following error message:

Warning:
Unable to create the remote backup: Transport error: Extension transport: ext://one-drive-backup/server/: cURL error 28: Operation timed out after 300000 milliseconds with 0 bytes received (see libcurl - Error Codes) for https://api.onedrive.com/rup/2a2060aa100c13d3/eyJSZXNvdXJjZUlEIjoiMkEyMDYwQUExMDBDMTNEMyE1OTQ2MiIs…….

On my second vServer run the backups daily to the same OneDrive account, without any problems:
Ubuntu 22.04.3 LTS and
Plesk Obsidian v18.0.54_build1800230824.08
Microsoft OneDrive Backup 1.2.0-65

Plugin Version:
Microsoft OneDrive Backup 1.2.0-65

What have I tried so far?:
- Logging out of my OneDrive account within the extension.
- Uninstalling and reinstalling the extension in Plesk

After uninstalling and reinstalling the extension in Plesk, I can log back into OneDrive and the first backup, right after logging in runs cleanly. When the backup is supposed to start automatically at night, it no longer works and I get the error message as seen above.

Unfortunately, I do not get on here and hope for your help!

 

[Kaspar]

The timeout error makes me think that there might be a firewall or networking issue. Either on your server, the datacenter or at Microsoft. Can you ping api.onedrive.com from your server? And if ping is successful can you manually cURL https://api.onedrive.com without error?

 

[httPete]

Hey @Kaspar thanks for your reply. I can ping api.onedrive.com from the server. Can you please explain me what i have todo here?:

And if ping is successful can you manually cURL https://api.onedrive.com without error?

 

[Kaspar]

Via SSH run curl -I https://api.onedrive.com. First line from response should be something like:

HTTP/1.1 405 Method Not Allowed

That means your server can connect to https://api.onedrive.com. Which is good. If it does, try starting the manually backup again. If the backup fails again, run the CURL command again immediately after backup failure. See if you still get the same response output from the CURL command.

 

[httPete]

Okay, thanks for your help. I have now taken the URL from my error message, which is displayed in Plesk.

I got the following error message:
curl: (3) [globbing] bad range in column 2

Then I changed the command to:

curl -I -g https://api.onedrive.com

(I shortened the full URL a bit here for the forum).

After that I get the following error message:
curl: (1) Protocol "[URL=https" not supported or disabled in libcurl

 

[Kaspar]

Did you ran the cURL command as root? If not, try running it again with root privileges.

 

[httPete]

Hey, unfortunately the link is always shortened by the forum (shortcodes)...

I have of course inserted the link in the command as it is also in the error message in Plesk.

Yes I ran it as root.

 

[httPete]

curl -I Microsoft OneDrive Dev Center | APIs and app development

Hello, @Kaspar I saw that you have adjusted the command again? So now I ran the following again as root. I got the following output:

HTTP/2 405
content-length: 103
content-type: application/json; charset=utf-8
p3p: CP="BUS CUR CONo FIN IVDo ONL OUR PHY SAMo TELo"
x-msnserver: DB3PPF695B17F28
strict-transport-security: max-age=31536000; includeSubDomains
ms-cv: kneTq86n/0SimCIsODliXw.0
x-qosstats: {"ApiId":0,"ResultType":2,"SourcePropertyId":0,"TargetPropertyId":42}
x-throwsite: 4d6d.dbc6
x-asmversion: UNKNOWN; 19.1258.922.2014
x-cache: CONFIG_NOCACHE
x-msedge-ref: Ref A: A303C74FA92B4E0781A34E7BE874021B Ref B: FRAEDGE1922 Ref C: 2023-10-06T09:10:21Z
date: Fri, 06 Oct 2023 09:10:20 GMT

 

[Kaspar]

The cURL output look fine. So both the Ping and cUrl commands you ran indicate that your server is able to make a connection to OneDrive. That's positive, but it unclear why backups fail with a cUrl timeout error. To troubleshoot this a bit further I would suggest running a backup manually from Plesk that is stored to OneDrive. If the backup fails again with the same timeout error, run the cUrl command via SSH again directly after failure. It would be interesting to see if the cUrl command will timeout too. My thinking is that maybe connections to OneDrive are temporary blocked (for whatever reason). But it could be something completely different too. I am not sure, as a timeout error is rather general.

 

[httPete]

sorry for the late feedback, but I was in bed with a cold for a few days. Here is some new information about my observations and tests:

On my Plesk server, I have set up a backup job via the Plesk Backup Manager that creates a backup every day at 1:30 both locally and on the remote server "OneDrive".

I have noticed that for some time now no incremental backups are created locally. Currently only full backups are created, although I have set up incremental backups in the Backup Manager. My fear is that due to the OneDrive error (cURL timeout), only full backups are created locally, since a current full backup is missing on OneDrive.

Could this be the reason that locally only full backups are created?

As a test, I specified the same OneDrive account as the backup destination directly in a subscription and created a backup. After the initial setup of the backup in the subscription, the backup ran cleanly. After that, I restarted several backups, each as a full backup:

- Configuration only
- Emails only
- Only database and data / website
- Config, E-Mails, Databases and Wesite (today)

These all went through cleanly.

Since tonight - until now, the OneDrive backup of the entire server is running and has also created data in OneDrive. However, "only" about 1 GB (each 100mb tar files) of about 23 GB... Which seems extremely slow to me.

Currently, I can observe that data is created in OneDrive, although it takes an incredibly long time.

Somehow, in my opinion, it does not seem to run quite correctly yet?

Since my first post here, I haven't changed anything either.... The only thing I did is that I turned off "incremental backups" once in the backup job (in Backup Manager). Then I saved the backup job and turned incremental backups right back on without starting a backup first. So I only changed the config once and changed it back again. I thought that maybe the config for the backup job is saved again and so the problem can be solved that incremental backups are created again and not only full backups.

 

[Kaspar]

sorry for the late feedback, but I was in bed with a cold for a few days. [...]

I hope you're feeling a bit better now.

The issue you a re describing is too broad for me to troubleshoot and really should be looked at in more detail. I also have very limited experience with storing backups to OneDrive. So I am not really sure what causing the issues you are having. Unless other forum users have some suggestions, the best suggestion I can offer is to contact Plesk support. So they can investigate the issue on your server.

 

[httPete]

@Kaspar thank you for your reply. Tonight the backup did not run automatically at 1:30 am. I think the job was still running because when I looked at 10pm last night, the OneDrive backup was still running and not finished. So I think it was still running at 1:30am....

I then manually started an incremental backup to internal storage and OneDrive at 6am this morning. Both ran through without error message. But there were probably again updates for Plesk and Ubuntu, maybe that brought something?! I will watch it and otherwise write to Plesk support.

But anyway, thanks for participating in my thread!