I’ve installed Urbackup a few times and it just keeps getting better with each release.
My current problem is that server works fine on both Windows or Linux and the client is fine on Windows. On Linux the client simply does not appear to talk to the server.
I have tried this on various combinations of Ubuntu x64 and i386 as well as Windows Server 2003, XP and 7.
I’ve built from source with no errors, and my log contents are as follows (debug level logging):
The first time I start the urbackup client:
“07/19/12 14:12:25: ERROR: Error preparing Query [SELECT tvalue FROM misc WHERE tkey=‘db_version’]: no such table: misc
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE shadowcopies ADD tname TEXT;]: duplicate column name: tname
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE shadowcopies ADD orig_target TEXT;]: duplicate column name: orig_target
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE shadowcopies ADD filesrv INTEGER;]: duplicate column name: filesrv
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE journal_ids ADD index_done INTEGER;]: duplicate column name: index_done
07/19/12 14:12:53: ERROR: Maximum wait time for thread pool exceeded. Shutting down the hard way”
Any following time:
07/19/12 14:13:21: Loaded -fileserv- plugin
07/19/12 14:13:21: Loaded -fsimageplugin- plugin
07/19/12 14:13:21: Starting as client…
07/19/12 14:13:21: Started UrBackup…
07/19/12 14:13:21: urbackupserver: Server started up sucessfully!
07/19/12 14:13:21: Info: retval 0
07/19/12 14:13:21: Info: retval 0
07/19/12 14:13:21: Info: Window size=253952
07/19/12 14:13:21: Binding udp socket…
07/19/12 14:13:21: done.
07/19/12 14:13:21: Servername: --
07/19/12 14:13:21: Server started up sucessfully
07/19/12 14:13:21: UDP Thread startet
07/19/12 14:13:22: Looking for old Sessions… 0 sessions
07/19/12 14:13:26: Receiving UDP packet…
07/19/12 14:13:26: UDP: PING received… sending PONG
07/19/12 14:14:17: Receiving UDP packet…
07/19/12 14:14:17: UDP: PING received… sending PONG"
"“07/19/12 14:12:25: ERROR: Error preparing Query [SELECT tvalue FROM misc WHERE tkey=‘db_version’]: no such table: misc
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE shadowcopies ADD tname TEXT;]: duplicate column name: tname
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE shadowcopies ADD orig_target TEXT;]: duplicate column name: orig_target
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE shadowcopies ADD filesrv INTEGER;]: duplicate column name: filesrv
07/19/12 14:12:25: ERROR: Error preparing Query [ALTER TABLE journal_ids ADD index_done INTEGER;]: duplicate column name: index_done
07/19/12 14:12:53: ERROR: Maximum wait time for thread pool exceeded. Shutting down the hard way”
These lines are obviously a bit of an issue. I have preformed a clean install by deleting all urbackup associated files from the system and recompiling and they still appear on first startup.
Those error messages are kind of a red herring. It upgrades the internal database on the first startup. Early databases did not have the ‘db_version’ version field. Thus the error messages. I’ll try to bundle already upgraded databases in the future. Thanks for pointing this out!
With regards to the actual error. This
07/19/12 14:13:26: Receiving UDP packet…
07/19/12 14:13:26: UDP: PING received… sending PONG
means that the Client is receiving the packets from the server and sending back answers. The server should discover the client given:
*It actually receives the UDP packet from the client. You could test this via tcpdump/ngrep/wireshark. It’s port 35622 on the server.
*The clientname has some content (i.e. is not blank)
*There is no other client with the same name
Good to know about those errors. Thanks for the help, I’ve been doing some more work on this and have observed the following:
From my Ubuntu Client and Server I’ve dumped the following data:
Client
22:36:58.970477 IP 192.168.0.24.32880 > 255.255.255.255.35622: UDP, length 1
E…@.@.z…p.&. 0…
22:36:58.971002 IP 192.168.0.223.35622 > 192.168.0.24.32880: UDP, length 8
E…$…@.@…&.p…i.$kumiko
Server
22:36:59.000449 IP 192.168.0.24.32880 > 255.255.255.255.35622: UDP, length 1
E…@.@.z…p.&. …
And throwing a windows box into the mix which works (dumps taken from server):
22:29:20.000532 IP 192.168.0.24.32880 > 255.255.255.255.35622: UDP, length 1
E…@.@.z…p.&. …
22:29:20.001282 IP 192.168.0.5.35622 > 192.168.0.24.32880: UDP, length 5
.E.$Rei…&.p.
The windows box Rei works fine, the Ubuntu box is still not registering with the server.
It looks like the client is sending but it never makes it to the server for some reason, any ideas on this one? On a related note I also had to remove the server (rm everything related to urbackup) and start again after a reboot of the box the server kept complaining about libcurl.so.4 not being found and then proceeded to segfault. Not sure if that’s relevant. (Yes libcurl was in place and I rebuilt 7 from source just to be safe).
I don’t know what could be wrong. If iptables on the client does reject the UDP packets one would not see them in tcpdump right? And the server does not reject them, because Windows clients with the same IP work… Btw. does it work if the client is on the same Ubuntu as the server?
The libcurl error is weird, but should not influence this. If it keeps occuring you can also just delete liburlplugin.so . Then the server cannot send mail but does otherwise work.
Very strange this one. After reading your comments I continued dumping more packets and looking through them, for a bit more background the whole system runs across two Vmware ESX nodes. So I tried a few physical boxes and a few more virtual ones. Everyone worked fine, and strangely after I updated the release from 10.04 to 11.10 things magically now work. I’m guessing a broken package or the like somewhere.
I’m still suffering from the Segfault issue with the backup server. Again a Vmware virtual machine running Ubuntu 12.04, After a fresh compile it starts and runs fine. The curl messages have now stopped appearing but if I reboot something goes wrong and the server segfaults on every startup meaning I have to remove everything to do with the installation and do a fresh install. I’ll try and narrow this one down a bit more, is this anything you’ve seen before?
Good that it works now . Did not see that curl error before. But as said as last resort if you delete the liburbackupserver_urlplugin.so or disable the url plugin loading in start_urbackup_server it will work fine but not send mails.
After a fresh install on a new Ubuntu VM I’m happy to say this is all working great!
I finally tracked down most of my problems to the “open-vm-tools” package not being installed on certain nodes which caused problems with the vmxnet3 NIC.