Since the applications, data, and services on my Rackspace (this) server is becoming more and more important I decided to look into Rackspace’s Cloud Backup service. However, I ran into trouble during the setup of the service, and this article tries to describe how it was solved.
I followed their guide to Install the Agent. There is an AUR for the driveclient that could be used, but at the moment I’m not that keen on using AUR’s, and at the time of writing the version in the AUR is also a bit old.
After configuring and starting the driveclient as instructed I went on to the second step Create a Backup. The backup failed and I examined the log file, /var/log/driveclient.log that yielded the following:
[2013-08-26 10:27:24|...|WARN |Http|rax::http::NetworkError::Type rax::HttpClient::ExamineLayer4Errors(474)] HTTP: Connection timed out while waiting for https://snet-storage101.lon3.clouddrive.com/v1.0/MossoCloudFS_ ...
[2013-08-26 10:27:24|...|ERROR|root|rax::Backup::PerformBackup(202)] Backup: Unable to complete the backup (1486201). Details: 1: [rax::HttpResponse local::<unnamed>::CurlEasyPerformWrapper(CURL*, rax::HttpContext&): 239-rax::HttpResponse local::<unnamed>::CurlEasyPerformWrapper(CURL*, rax::HttpContext&)] Ex Code(2000): Could not perform an HTTP request. Timeout was reached (28)
[2013-08-26 10:27:24|...|ERROR|root|rax::Backup::MakeReport(332)] Failed to start backup: 1: [rax::HttpResponse local::<unnamed>::CurlEasyPerformWrapper(CURL*, rax::HttpContext&): 239-rax::HttpResponse local::<unnamed>::CurlEasyPerformWrapper(CURL*, rax::HttpContext&)] Ex Code(2000): Could not perform an HTTP request. Timeout was reached (28) Error during backup initialization.
This made me turn to the always excellent Customer Support at Rackspace, and it seemed that I were missing some routes. I issued the following commands to add the correct routes to the service net (eth1):
# ip route add 10.176/12 via 10.179.64.1
# ip route add 10.208/12 via 10.179.64.1</pre>
I restarted the backup from Rackspace’s Control Panel and it completed successfully.
But after rebooting my server the routes were ones again gone and the backup subsequently failed. At this point I started to examine the netctl profile for eth1 (/etc/netctl/eth1). The configuration file stated:
ROUTES=('10.208.0.0/255.240.0.0 via 10.179.64.1 10.176.0.0/255.240.0.0 via 10.179.64.1')
which is the netcfg configuratiion. Simply change ROUTES to Routes and, for readablity add ' to separate the instructions (netctl will fix this by itself if you don’t add the '):
Routes=('10.208.0.0/255.240.0.0 via 10.179.64.1' '10.176.0.0/255.240.0.0 via 10.179.64.1')
To finish up the installation of the driveclient I added the following to the service file /etc/systemd/system/driveclient.service:
[Unit]
Description=Rackspace backup agent
After=network.target
[Service]
ExecStart=/usr/local/bin/driveclient
[Install]
WantedBy=multi-user.target</pre>
Followed by enabling and starting the service:
# systemctl enable driveclient
# systemctl start driveclient
I’m told that the Rackspace team is now aware of the problem with the routes configuration for the eth1 interface and should not cause any problem when installing a new Arch Linux server.