I am trying to test couchbase enterprise edition 5.5
In my tests i am trying to do a full backup but it seems something happens in that process.
This is what i get :
2018-10-07 20:17:55,073: w0 no response for 30 seconds while there 1024 active streams
[ ] 0.1% (37151/estimated 34631186 msgs)
bucket: logs, msgs transferred…
: total | last | per sec
byte : 7103643 | 7103643 | 214181.1
done
Has anyone had this problem before ? Or do you have any ideea what it means ?
Are you using the command line tool to perform a backup? Could you show exactly what the command is that you’re typing? Do you get the same error on a smaller bucket?
I am using the following command : ./cbbackup couchbase://127.0.0.1:8091 /backups/10 -u SomeUser -p SomePassword -m full
I did not try on a small backet. But i think the error will not appear. On the big bucket sometimes i even reach 40%
2018-10-30 19:24:43,594: w0 no response for 30 seconds while there 1024 active streams
[######### ] 43.9% (15692540/estimated 35754538 msgs)
bucket: logs, msgs transferred…
: total | last | per sec
byte : 3440284582 | 3440284582 | 2707410.7
Is it possible that you are running cbbackup multiple times at once? If so, there is an outstanding issue around that. Loading...
If not, then we’ll need to create a new ticket. Could you please try it on a small bucket just to confirm that this only occurs for a large bucket? Could you also tell me what operating system you are using (and version information), as well as how many nodes are in your Couchbase cluster?
The error message means that cbbackup did not receive any data from the cluster. Would it be possible to run the command again with -vvv to increase the verbosity of the logging. It would also be worth checking the server side logs. The memcached.log will show what is happening with the cbbackup connection.
@ciprian Can you try this on Couchbase Server 6 please. I suspect you might be hitting MB-31481 where in rare cases the server does not send the snapshot end marker.