NetBalancer Bug in version 3.3.2

Home Page Forums Network Management Signal a BUG NetBalancer Bug in version 3.3.2

This topic contains 4 replies, has 0 voices, and was last updated by  danrwalker 3 years, 9 months ago.

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
    Posts
  • #44193

    danrwalker
    Member

    Hi,
    I have installed a copy of ZeroShell 3.3.2 on our local network and the NetBalancer only seems to work if you have a single router enabled. Enabling more than 1 route results in lots of failed page loads and connections for users.

    We have previously being useing version 2.0 rc2 without issue.

    I have rolled back to version 3.2.0 using the same profile and all appears to work well.

    Dan

    #53697

    Matze
    Participant

    I also have problems with the NetBalancer in the new version, see this entry:
    https://www.zeroshell.org/forum/viewtopic.php?t=1072

    #53698

    erick06fr
    Member

    Same problem with 3.3.2 : net balancer will not bring up again PPOE interface after a drop, and stay on “failed” mode and “alert” until ZS reboot.

    No solution. Rolling back again to 3.2.1, wich is fully functionnal, in spite of security holes.

    EDIT 19h22 : Made a test with ZS 3.3.1 : netbalancer is OK with PPPOE failure recovery.

    so if resume :

    ZS 3.2.1 : Netbalancer OK
    ZS 3.3.1 : Netbalancer OK
    ZS 3.3.2 : Netbalancer KO 😥

    #53699

    erick06fr
    Member

    Very strange beahviour on netbalancer.

    this evening, after the weekly pppoe connection operator drop, ZS 3.3.1 was unable to change the netbalancer to UP status for this interface. Pppoe is connected, as I can see in the log, but still found down by ZS in the netbalancer and considered as Failed

    Just clicked on the “save” button of the netbalancer, and the connection is immediatly put in the UP status..

    I’ve immediately tested a manual forced failure by disconnecting the DSL phone line of the pppoe modem. ZS netbalancer rightly change to FAULT again on pppoe. Plugin on the phone line again, and after one minute, the netbalancer pppoe was online again, perfectly.

    So, netbalancer seems now to be KO on operator-side connection drop, and OK on end-user-side phone line OFF/ON test.. This is not logic !

    This is very frustrating, as the netbalancer runs well from many years, and after testing 3.3.2, in which version it is totaly broken, now it seems to be broken in previous releases too.. May be the 3.3.2 corrupts the database net balancer configuration in some way ?

    I will try if I have any time to restore a databse backup.
    I need also to test net balancer a full week with the rock-solid old ZS 3.2.1

    #53700

    erick06fr
    Member

    After one whole week of test of Netbalancer with ZS 3.3.1 and 3.3.2, it’s now obvious that there is a big bug in these versions as soon as one member of the Netbalancer pool is a PPPOE link.

    As said previvously, NetBalancer won’t bring up Online the PPPOE link after a fault (typically the regular operator drop to change public IP). It’s necessary to manually click on the “save” button of the NetBalancer setup, at each time the link went down, to bring the link up again.

    Reverting back to ZS 3.2.1 with the same configuration file solves the issue.
    Hope that “the boss” can fix this problem, as NetBalancer users with PPPOE internet providers cannot run ZS in last versions including security fixes.

    #53701

    Lightyear
    Member

    Same behavior here. Going back to 3.2.1
    Thank you erick06fr for your thorough work! Saved me a lot of time!

Viewing 6 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic.