Using a DNS-320, current firmware 2.02 with 2x 2TB Seagate ST2000DL003 in RAID1
Trying to use remote backup (Rsync) to take a backup copy of files on a company Linux server onto the NAS for off-site storage. NAS and Linux server (Fedora 16) on same LAN (100base-T). There's about 1.5TB of data to be backed up altogether...
1. Got Rsync daemon running on Linux box and a suitable rsyncd.conf file that gives me read access to the folders I want to back up. I have created several "modules" in that file so I can back up each folder tree separately.
2. I can create the manual backup job on the NAS. I have the following observations as I create the job:
(a) I choose the kind of backup job as "NAS to Linux" and "Remote to Local"
(b) I enter the job's name and the folder on the NAS to store the backup.
(c) I enter the Linux server's IP address and the user name/password to use. I don't enable incremental backup support as I'm just after a complete snapshot of what is on the Linux server...
But when I click "Next" there can be a very lengthy "hang" - the status may show as "Testing". This hang time seems to be dependent upon the number of files/folders in the folder tree in the Linux server; in an extreme case with one of the larger folder trees it exceeded the login inactivity timeout on the NAS and it logged me off before I could complete the wizard!! I lengthened the inactivity period to 15 minutes...
(d) I competed the wizard, with the job as "manual" and did not choose to launch it immediately.
(e) Upon clicking the "Backup now" icon in the job's line, the job started; a percentage showed, then after a few minutes it showed as "Finished" But no files were (yet) visible in the NAS!!
However, I could see by refreshing the statistics on disk usage, and by looking at Ethernet activity lights on the switch and the hard disk activity lights on the NAS, I could plainly see that the job had definitely _not_ finished!!
So I left it running in the belief tta the files were being downloaded into a temporary hidden folder and would "magically appear" when the job really did complete...
However, a couple of hours later, it terminated abnormally:
i. There was nothing in the NAS logs or job status to show that anything had failed
ii. The rsyncd.log file on the Linux server showed a "connection reset by peer" error at the time of the failure
iii. About 100GB of disk space had been consumed on the NAS; none of this was visible in any user accessible folders, nor did there seem to be any way to flush it other than reformatting the disks!!
I am trying again - splitting the backup into smaller jobs of 50GB or less, but wonder if:
(a) Is there any known issue with the remote backup - such as limit on backup size or number of files/folders that can be handled by one backup job?
(b) If the backup does fail - is there a way to flush the hidden files - or will they automatically be flushed the next time I run a backup job?
Thanks,
Richard