Hi,
I had some trouble to finish the client install on windows 7. The installer was not able to start the service. After some try and fail attempts I found that the “api-ms-win-crt-runtime-l1-1.0.dll” was not installed and the client depends on that.
After I’ve installed kb2999226 from microsoft the service could be started.
Could you add the dll to the installer?
Regards,
BerndG
PS: Thank you for this piece of software.
I experienced similar problem. I will try KB2999226 and see if this is a common issue.
Thanks for the investigation but this is a OS update which should be (automatically) installed by Windows update. My internet research shows me that other installers look if this patch is installed and then cancel with a error message to the KB if not.
You’ll not only have problems with UrBackup if this OS patch is missing.
UrBackup did a tidy because I was running out of space. It removed a bunch of incremental image backups of the C Drive but left the folders. I now have a bunch of these:
It’s also left all the SYSVOL & ESP images intact, which I imagine are useless:
and…
…also, does it ever tidy up file backup folders. I know they’re all hardlinked so doesn’t take up any space but will I eventually be left with 100’s of dated file backup folders?
It only truncates the files to zero bytes after it failed to delete them. Do you have a virus scanner or something similar which may block file deletion?
Nope. It’s just a Windows 10 PC, no anti-virus or anything like that. No fancy security setups or users.
Is it safe to delete them?
Windows 10 comes with anti-virus built-in and enabled by default (Windows Defender). Was it disabled?
Yep, my mistake. No, it wasn’t. I’ve never seen it lock 14 separate folders for that length of time while something like Urbackup tries to remove them though and it certainly wouldn’t purposely block folders from being deleted would it?
@uroni Is it safe to delete these folders now? And why do I also have 28 corresponding SYSVOL and ESP folders that still have vhdz’s taking up space for each of the empty C folders? Is that correct as part of the tidy or do they need to be removed too?
Thanks for your time.
Theoretically the SYSVOL and ESP volumes should have been deleted. I’ll have a look at that. Maybe it will have the errenous behaviour again if you run cleanup.bat such that it cleans something and can post the debug log ?
I’ll give it a go and post the logs if there’s a problem. Thanks
Hi,
Debian 8 / UrBackup Server (2.0.21) / btrfs-progs v 4.4 / Windows 2012R2 clients (2.0.19)
This morning we have few backups (files + images) stalled and a high load on the urbackup server (up to 20 !)
And iotop shows very high rates :
Clicking on the “Stop” button doesn’t stop backups,
Image backups log output :
File backups log output :
Regards,
Getting “Constructing of filelist of “PCNAME” faild: error - stopped indexing 2” again since update to 2.0.19 client…
15 PCs are failing backups which did run fine.
Those are expected if you changed the backup folders during indexing. Could this be the case?
Nope… No Settings changed at all…
Only updated the clients.
Hi Martin. Just to let you know that the error didn’t reoccur when I ran cleanup.bat so no entries in the log for you this time!
Also, running remove_unknown.bat removed all those SYSVOL and ESP folders so I’m all good for now. I’ll see what happens the next time i run out of space.
Do you have a client or server debug log or can you create a memory dump of the client?
Also interesting: It is a resumed backup. Why did the previous incremental file backup fail? (And preferably a full log of that as well)
@tarakesh Do you have, or could you produce a client debug log? Thanks!
Also interesting: Did you only upgrade from 2.0.18 to 2.0.19 and what operating system(s) do the clients have?
Debian 8.4 / Server 2.0.21 / Clients 2.0.19 / BTRFS v4.4 / Kernel 4.5.0
Tonight i decide t set the “max simultaneous backups” value from 12 to 4 in order to decrease the server load and to see if it helps to achieve backups more quickly. We’re backuping 12 Windows 2012 R2 Servers and maybe it’s too much in one time.
All of the clients have the same scheduling, i.e “Mon-Thu/22-24;Tue-Fri/0-6”.
But this parameter doesn’t seem to be respected, server launch more than 4 backups :
How can we improve this ?
Regards,
I upgraded to 2.0.21 today. So far so good!
I did notice that when I paused a user’s initial backup when they were indexing, the user’s status continuously shows “indexing” on the server side. I would think it should complete the index and quit or just show a status of “paused”. Also, can the admin unpause a user’s backup (does a request to start a full or incremental undo the pause)?
