WAN Disconnects during load of firewall_aliases_edit and interfaces_wan pages



  • Hi,

    I'm using the 1.2 RC3 release, and experiencing some weird behaviour when opening the interfaces_wan.php and firewall_aliases_edit.php. When I'm trying to open these pages, the page won't get loaded and the WAN connection disconnects (my pings won't go through for a while, giving request timed out). After cancelling the page load (using escape) and waiting for a few seconds the ping gives response again. Only the WAN connection disconnects, I cannot find anything using dmesg or syslog. My LAN and OPT interface are responding during this time. The machine is not using much cpu and / or memory.

    I've tried this for several times and it is consequent and reproducable, but I cannot get a clue about the cause of this. Any ideas?

    Remco



  • @Remco:

    Hi,

    I'm using the 1.2 RC3 release, and experiencing some weird behaviour when opening the interfaces_wan.php and firewall_aliases_edit.php. When I'm trying to open these pages, the page won't get loaded and the WAN connection disconnects (my pings won't go through for a while, giving request timed out). After cancelling the page load (using escape) and waiting for a few seconds the ping gives response again. Only the WAN connection disconnects, I cannot find anything using dmesg or syslog. My LAN and OPT interface are responding during this time. The machine is not using much cpu and / or memory.

    I've tried this for several times and it is consequent and reproducable, but I cannot get a clue about the cause of this. Any ideas?

    Remco

    It appears that I'm also getting watchdog timeouts when having little load (1Mbps), I'm investigating this issue and let you know if I find out something.



  • When pinging from the interface which is not responsive I'm getting the following error:

    ping: sendto: No buffer space available
    

    with netstat -m

    1513/677/2190 mbufs in use (current/cache/total)
    661/621/1282/0 mbuf clusters in use (current/cache/total/max)
    643/125 mbuf+clusters out of packet secondary zone in use (current/cache)
    0/0/0/0 4k (page size) jumbo clusters in use (current/cache/total/max)
    0/0/0/0 9k jumbo clusters in use (current/cache/total/max)
    0/0/0/0 16k jumbo clusters in use (current/cache/total/max)
    1700K/1411K/3111K bytes allocated to network (current/cache/total)
    0/0/0 requests for mbufs denied (mbufs/clusters/mbuf+clusters)
    0/0/0 requests for jumbo clusters denied (4k/9k/16k)
    0/5/4368 sfbufs in use (current/peak/max)
    0 requests for sfbufs denied
    0 requests for sfbufs delayed
    0 requests for I/O initiated by sendfile
    0 calls to protocol drain routines
    
    


  • sysctl -a

    
    ITEM            SIZE     LIMIT     USED    FREE  REQUESTS
    mbuf_jumbo_1:  16384,        0,      0,      0,        0
    mbuf_jumbo_9:   9216,        0,      0,      0,        0
    mbuf_jumbo_p:   4096,        0,      0,      0,        0
    mbuf_cluster:   2048,        0,    786,    496,     9929
    mbuf:            256,        0,   1638,    552,   188280
    mbuf_packet:     256,        0,   2051,    139,    40070
    
    


  • After some googling it seems to be the same issue as http://www.freebsd.org/cgi/query-pr.cgi?pr=78801, but at my configuration the kern.ipc.nmbclusters="0" isn't working and the issue arises in a configuration with low traffic.

    An other box we have deployed seems to have no problems with the 6.2RC2 (and before) releases. Any idea's?



  • Finally found the cause of the strange behaviour. It appears I had also a 1:1 NAT mapping and this conflicts with the Virtual IP (proxy arp). After looking after the firewall_virtual_ip_edit.php it appears it is checking the wrong field ($_POST['ipaddr']  instead of $_POST['subnet']) for this kind of errors and proceeds adding the virtual ip.


Log in to reply