I'm a frequent user of veeam agent for windows and just started to deploy the veeam agent on two linux machines. Everything installed fine and after doing one more necessary step (https://helpcenter.veeam.com/docs/agent ... ccess.html) my user can run veeam.
I did all the onboarding things and created a new file based backup job with target to a local attached usb storage (tested with a usb2 stick and usb3 hd). The two devices I tested both have the OS on a eMMC storage. Veeam get's it right and starts even the backup (mmcblkX,..) but it always fails to complete the job. It looks like the maximum amount of data it reaches for backup is around 10GB (which is around 60% on one of the machines).
I tried many times but never suceded
Both machines I tested run the amd64 version and using the deb packages provided by veeam. OS I tested is Debian 9 and Linux Mint 19.
22:44:35 Creating volume snapshot 00:00:02 ^
22:44:37 Starting full backup to Repository_1 │
22:44:38 Backed up mmcblk1boot1 4 MB at 5 MB/s 00:00:01 │
22:44:39 Backed up mmcblk1boot0 4 MB at 10.3 MB/s 00:00:01 │
22:44:40 [error] Backing up mmcblk1rpmb 00:00:00 │
22:44:47 [error] Failed to perform backup │
22:44:47 [error] Input/output error │
22:44:47 [error] POSIX: Failed to read data from the file [/dev/veeamimage2] │
22:44:47 [error] Failed to upload disk │
22:44:47 [error] Agent failed to process method {DataTransfer.SyncDisk} │
22:44:47 [error] Exception from server: Input/output error │
22:44:47 [error] POSIX: Failed to read data from the file [/dev/veeamimage2] │
22:44:47 [error] Unable to retrieve next block transmission command. Number of already processed blocks: [0]
In general, the Veeam Agent for Linux does not care about whether it is running on SSD, eMMC, spinning disk or whatever. So I would not focus on that.
As far as I see, there are no logs attached to the case. Support cannot help without logs. Please do also understand, that the forum does not replace support.
However, since eMMC is a slightly different format of memory cards, this will need a clarification from our development team.
Whose quote is that? AFAIK eMMC is not really different from any other multi-media card except for the fact that it is embedded. By the way, what is the file-system on that eMMC?
Are there any other storages on the machines? My guess is that the issue might related to a snapshot being located on the same eMMC which might cause the IO errors as some eMMCs operate in half-duplex mode, and that is something what our snapshot algorithm might have not taken into account.
It seems that the support case is missing logs, would you share them please? You can upload them to some sharing service and PM me the link.
The error seams related to the device "/dev/mmcblk1rpmb" which is the "Replay Protected Memory Block (RPMB) subsystem" and not readable.
I good this very useful tip from the veeam support:
As a way to work this around I'd suggest trying to skip this partition.
To do so, please follow these steps:
1) Make sure no backup jobs are ruunning ion the agent
2) Stop the "veeamservice" service
3) Open /etc/veeam/veeam.ini, and add the "ignoreDevices= /dev/mmcblk1rpmb" at the end of the [backup] section of the file
4) Start the veeamservice and try to take a backup again.
Ok, I got ahold of the logs and passed it to our dev team, stay tuned.
Open /etc/veeam/veeam.ini, and add the "ignoreDevices= /dev/mmcblk1rpmb" at the end of the [backup] section of the file
This thing will make Veeam to skip the specified devices. That is, it will pretend that mmcblk1rpmb does not exist, which, I believe, is not exactly what you were looking for since it is your only storage.
