Hi, 
Its most probably a situation where your system runs out of resources. 

One thing to watch may be to use `db._explain(query)` in arangosh or the 
webinterface and find out whether all of your queries properly use indices.

Another thing that may is that i.e. you're targeted by a botnet, so traffic 
increases above the expected limits.

One thing to fine tune could be the maximum number of open files:
https://www.cyberciti.biz/faq/linux-increase-the-maximum-number-of-open-files/

In doubt i'd put in a rater limit to the frontend like described here: 
https://debian-administration.org/article/187/Using_iptables_to_rate-limit_incoming_connections

But yes, having a look at the number of open connections is definitely one 
thing to do once this reoccurs.


Regarding your WAL-log issues, you can force arangodb to ignore them by 
using `--wal.ignore-recovery-errors` but you should dump & re-import your 
data afterwards. 

Cheers, 
Willi

On Friday, April 7, 2017 at 6:21:37 PM UTC+2, Georgios Kafataridis wrote:
>
>
>
>
>
>
> OK, I use the php driver and I was already using it the keep-alive flag.
>
>     'class' => 'app\components\arangodb\Connection',
>     'endpoint' => 'tcp://10.69.198.2:8529', //internal database server
>     'username' => '',
>     'password' => '',
>     'database' => '',
>     'persistence' => app\components\arangodb\Connection::
> PERSISTENCE_KEEPALIVE,
>
> If this happens again I should look for  hanged connections ? 
> So it's not a problem of arangodb but problem of the driiver? 
>
> Restarting arangodb3 service seems to close all connections, but doing so 
> has left me with corrupted wal journals.
>

-- 
You received this message because you are subscribed to the Google Groups 
"ArangoDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to