Snap Deploy 2.0 - DHCP problems

Discussion in 'Other Acronis Products' started by SOE, Jul 20, 2006.

Thread Status:
Not open for further replies.
  1. SOE

    SOE Registered Member

    Joined:
    Jul 11, 2006
    Posts:
    8
    I have been trying to get Snap Deploy 2.0 to work with several
    Dell Optiplex GX520's. I have managed to create a master image of
    one of the GX520's and have saved it to the server. But when I go to
    boot up from the CD or PXE to deploy the image the deploy agent
    doesn't get a DHCP assigned IP address. So I have to enter all the
    details manually. The weird thing is When I PXE boot, my NIC needs to
    get a DHCP assigned IP address in order to find the PXE server and run
    the Snap Deploy Agent (which of course it does).
    If my NIC card can get an IP I'm wondering why the Deploy Agent can't?

    All PC’s are networked on a large domain with internal DNS, DHCP, and
    WINS servers running and all are detected fine within windows.
    The thing is, last years Dell Optiplex GX280 works fine with Snap
    Deploy. It gets an IP address from the DHCP server and
    automatically recognises the Deploy servers IP address.
    So my natural conclusion is hardware support for this late model Dell.

    I have a Broadcom Tigon 3 Gigabit Card (but I have 50 PC's with them),
    and are a motherboard intergrated item like so many these days.
    The motherboard is a Foxconn LS-36.

    I realise that adding support for every card would be a huge undertaking
    so is there some way we can add drivers for the deploy agent?
    Anyone else having NIC card support problems?

    Any help would be much appreciated.
     
  2. Acronis Support

    Acronis Support Acronis Support Staff

    Joined:
    Apr 28, 2004
    Posts:
    25,885
    Hello SOE,

    Thank you for using Acronis Snap Deploy

    Could you please boot your computer from Acronis Snap Deploy Bootable Agent and press F11 key after the "Starting Acronis Loader..." message appears and you get to the selection screen of the program? After you get the "Linux Kernel Settings" prompt, please remove the word "quiet", click on the "OK" button and wait for # prompt to appear. Please issue the following commands:

    asamba dhcp -n
    asamba masters
    cat /proc/net/dev


    Please write down the output information after issuing each of the commands, or use a digital camera to make shots of the output and post it in this thread. Please note that you can use Shift+PgUp and Shift+PgDn keys to scroll the screen manually.

    Thank you.
     
  3. SOE

    SOE Registered Member

    Joined:
    Jul 11, 2006
    Posts:
    8
    Attached are the results of the commands in screencapture form.
    The first two commands fail.
    Repeating them seem to correct this somehow though.
     

    Attached Files:

  4. SOE

    SOE Registered Member

    Joined:
    Jul 11, 2006
    Posts:
    8
    repeating the first two commands show this:
    Code:
    [B]asamba dhcp -n[/B]
    info, udhcp client (v) started
    SIOCDELRT: unknown error 3
    debug, Sending discover...
    debug, Sending select for 130.217.41.188
    info, Lease of 130.217.41.188 obtained,  Lease time 
    SIOCDELRT: unknown error 3
    Repeating this command finally gets the deploy agent to see the DHCP server and get an IP from it.


    Code:
    [B]asamba masters[/B]
    130.217.140.41 SUS-SERVER
    It can now also see our SUS server on the domain as well which is one of the master browsers.
     
  5. jbillerbeck

    jbillerbeck Registered Member

    Joined:
    Jul 30, 2006
    Posts:
    1
    I have the exact same problem with our brand new GX620's. It has to be a problem with the Bios or the NIC cards (Broadcom's). However I do believe that with TENS of Thousands of these Dell PC's out there that Acronis really should look at fxing the client. It looks as though it justs needs a couple of tries for DHCP to take. ( I tried the samba command line stuff and it worked everytime on the second try) It will be a real pain if we have to go to every PC that is being connected for deployment and re-enter lines of code etc. With PXE or CD boots it should be possible to start the machine , point the client and walk away. What good is DHCP otherwise ? Anyone de-compile rhe acronis client make changes and recompile ? It should not be real difficult and the code that needs to be added seems pretty simple. :cautious:
     
  6. SOE

    SOE Registered Member

    Joined:
    Jul 11, 2006
    Posts:
    8
    It's been about a week since I last posted. Just wondering if there's some sort of solution to wait for......?
     
  7. sjucht

    sjucht Registered Member

    Joined:
    Aug 1, 2006
    Posts:
    1
    Are you using cisco switches? We were and we had to set the ports used for imaging to spanning-tree portfast trunking in order to get DHCP to work.

    Just something to look at!

    S
     
  8. SOE

    SOE Registered Member

    Joined:
    Jul 11, 2006
    Posts:
    8
    Actually yes we do have Cisco switches. I passed on the above comment to a couple of our network engineers who told me its was more to do with the "Spanning Tree" (switch/router mumbo jumbo). Suffice to say that their recommendation to increase the timeout of the network on the target PC from 60secs to 120secs would solve the problem.

    This actually seems logical to me as in a previous post I wasn't able to get an address the first time around when performing the Samba DHCP -n command. But repeating this command would yeild a DHCP address.

    The only thing is the Dell bios that I'm dealing with won't allow for that level of adjustment.

    So now I know what I have to do, I just don't know how. ARRGGHH!

    Anyone know a way of adjusting the network timeout threshold on a PC that won't let you change it at a bios level? (Baring in mind that Windows Xp manipulation will not affect a PXE boot)
    Is there a command that i can run to increase the timout?
    Is there and option in snap deploy that will let me increase the time out?
    If not, will there be?

    Hmmmmmm.....
     
  9. bswaby

    bswaby Registered Member

    Joined:
    Aug 9, 2006
    Posts:
    7
    If they are not willing to turn portfast on then the only real option that you have is to put a cheap switch or hub between the computer and Cisco switch. This will keep the port up and the cheap switch will allow your computer to pull the dhcp address from your server.

    Just for the record I have been able to image from a gx270 to the gx620 with no issues. The gx620 had bios level A01 prior to me imaging it. Once it was imaged I updated it to A08, which i don't know if it has a problem with that version or not.

    Ben
     
  10. hittman

    hittman Registered Member

    Joined:
    Aug 10, 2006
    Posts:
    1
    o_O Just wondering if someone could tell me something. Is it possible that a pc or network printer could to send out BPDU? The reason I ask is that I am being accused of putting a switch on our network and disabling the port (which they say needs to be 'partitioned'). Which I guess is an automatic process. Here is what I have from the network guys.

    Jul 11 19:02:46.946: %SPANTREE-SP-2-BLOCK_BPDUGUARD: Received BPDU on port GigabitEthernet8/6 with BPDU Guard enabled. Disabling port.
    Jul 11 19:02:46.946: %PM-SP-4-ERR_DISABLE: bpduguard error detected on Gi8/6, putting Gi8/6 in err-disable state

    From this they say that they can tell it was a switch and that only switches do this, not pcs and printers.

    Also, can anyone tell me what 'Partitioning' a port means? Or are they blowing smoke? They are known to make things up and have a very hard time saying 'I don't know why', but very easy time placing blame.

    Thanks
     
  11. SOE

    SOE Registered Member

    Joined:
    Jul 11, 2006
    Posts:
    8
    bswaby:
    Do you mean put a switch/hub between the server and the cisco switch or client pc and cisco switch?

    I have tried both ways. Neither seems to make a difference. I have a small 5-port D-link Switch which I had place in-between the Sever and the Cisco switch (the little 5 port is now a sort of go-between for the server network connection). I then proceeded to do a net boot on the Client PC (which was able to grab a DHCP address) and find the PXE server and boot the agent. When its the agents turn to get a DHCP address it always fails. Same result with the Dlink switch as a go-between for the client pc to Cisco switch.

    So no change really.
     
  12. bswaby

    bswaby Registered Member

    Joined:
    Aug 9, 2006
    Posts:
    7
    It's really going to be between the client and Cisco switch. The thing that I may have missed in the original posting is that you are getting an ip address from the PXE server with or without the d-link switch in between. What puzzles me and makes me lean to a Acronis problem is that fact that you used PXE to boot and get the image to start Acronis. Because you had to have a DHCP address to do all of that.

    This most likely doesn't matter, but what PXE server are you using? Also, have you contacted Acronis support by telephone? They are a little hard to get ahold of but maybe phone support could walk you through a few things

    FYI: I'm not saying this is your problem, but I have had a lot of problems out of D-Link products in the past.

    Ben
     
  13. nilrod

    nilrod Registered Member

    Joined:
    Aug 24, 2006
    Posts:
    1
    Hello.

    I just want to add that I am experiencing the exact same problem. And I'm using only IBM computers; no Dells here.
     
Thread Status:
Not open for further replies.