Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Advertise on LowEndTalk.com
MyCustomHosting - VM not booting / customers fault
New on LowEndTalk? Please read our 'Community Rules' by clicking on it in the right menu!

MyCustomHosting - VM not booting / customers fault

tr1ckytr1cky Member

So apparently I have a problem with MyCustomHosting.
Let me summarize the process:

  1. I get an email from MCH that they will do node reboots.
  2. An hour later I get another email that node reboots are now finished.
  3. I notice my VM isn't up and I try to boot it, but it doesn't.
  4. I open a ticket explaining the issue.
  5. Phil tells me it's an issue on my side and he can't do anything, because the vm is unmanaged. He suggests a reinstall to fix the issue.

So apparently the provider updates the node and afterwards some VMs (a friends VM doesn't boot either) don't boot anymore, I guess that's totally the customers fault.
I never had such problems with any other provider I've been with and Phil isn't even willing to look into the issue, instead he tells me I have to solve this myself (allthough VNC isn't working either, so how should I?)

tsdns.io - free, redundant, DDoS-protected TSDNS

«13

Comments

  • did you get console access? Because console is routed through master node so you might see the issue where your vm is clogged from booting.
    But 1 hour for node restart seems to be a bit of overkill

  • @tr1cky said:
    So apparently I have a problem with MyCustomHosting.
    Let me summarize the process:

    1. I get an email from MCH that they will do node reboots.
    2. An hour later I get another email that node reboots are now finished.
    3. I notice my VM isn't up and I try to boot it, but it doesn't.
    4. I open a ticket explaining the issue.
    5. Phil tells me it's an issue on my side and he can't do anything, because the vm is unmanaged. He suggests a reinstall to fix the issue.

    So apparently the provider updates the node and afterwards some VMs (a friends VM doesn't boot either) don't boot anymore, I guess that's totally the customers fault.
    I never had such problems with any other provider I've been with and Phil isn't even willing to look into the issue, instead he tells me I have to solve this myself (allthough VNC isn't working either, so how should I?)

    If VNC isn't working it might suggest something beyond your control. Ensure that boot order etc are all fine. You might want to ask MCH if there are any errors with the domain (assuming KVM) during startup. Was VNC working for you before?

  • tr1ckytr1cky Member
    edited April 2014

    Okay, apparently after changing java security settings, vnc works.
    It now says:

    "Booting from Floppy... Boot failed: could not read the boot disk

    No bootable device."

    Boot device is hard disk.

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    You changed the disk driver.

    http://pastebin.com/YhDG7tdd

    I can't handle all the variables you are asking me to for an unmanaged service. I made sure to have the node online. Sorry it seems a form completion plugin likely or etc changed the disk driver on your VM.

    I was more than willing to look into the issue for you. I rebooted your VM. Checked VNC was working and even sent you a screenshot to show you what I was seeing. Connected to your VM's VNC.

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • VirtovoVirtovo Member
    edited April 2014

    @tr1cky said:
    Okay, apparently after changing java security settings, vnc works.
    It now says:

    "Booting from Floppy... Boot failed: could not read the boot disk

    No bootable device."

    Boot device is hard disk.

    Ask MCH to see if the image is being found when the domain starts. If using SolusVM they can do this quite easily using the debug boot option otherwise you can get the same info starting the domain from the command line.

  • @MCHPhil Here we go again....:)

  • @MCHPhil said:
    You changed the disk driver.

    http://pastebin.com/YhDG7tdd

    I can't handle all the variables you are asking me to for an unmanaged service. I made sure to have the node online. Sorry it seems a form completion plugin likely or etc changed the disk driver on your VM.

    Yes, I did because I wanted to make sure every possible thing on my side is done before I ask you.

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    The error is because you changed the disk driver. I don't know how that is my fault? That is VERY common error if you change it after the OS is installed. As I suggested, reinstall the VM.

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • @MCHPhil said:
    The error is because you changed the disk driver. I don't know how that is my fault? That is VERY common error if you change it after the OS is installed.

    It wasn't booting, so I changed the disk driver to see if it does then.

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • irmirm Member
    edited April 2014

    Yeah I was having issues with my VM after the node reboot also.

    It was saying "Online" in SolusVM yet couldn't ping it or connect to it through SolusVM VNC or through SSH. Eventually after enough reboot/boot attempts in the SolusVM control panel and a "Reconfigure Network" attempt or two in the SolusVM control panel, I was able to access my VM via VNC and soon after via SSH which I guess is a success but leaves me wondering why it took so many attempts to get it going :l

  • MCHPhilMCHPhil Member, Provider

    @irm said:
    Yeah I was having issues with my VM after the node reboot also.

    It was saying "Online" in SolusVM yet couldn't ping it or connect to it through SolusVM VNC or through SSH. Eventually after enough reboot/boot attempts in the SolusVM control panel and a "Reconfigure Network" attempt or two in the SolusVM control panel, I was able to access my VM via VNC and soon after via SSH which I guess is a success but leaves me wondering why it took so many attempts to get it going :l

    It is well known it can take a while to do a node reboot. As I stated in the email sent before this happened. Starting tons of VM's at once takes time. It isn't a 10 minute process. Likely your VM was still booting.

    Thanked by 1CFarence
    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • @tr1cky said:

    1. I get an email from MCH that they will do node reboots.
    2. An hour later I get another email that node reboots are now finished.
    3. I notice my VM isn't up and I try to boot it, but it doesn't.
    4. I open a ticket explaining the issue.
    5. Phil tells me it's an issue on my side and he can't do anything, because the vm is unmanaged. He suggests a reinstall to fix the issue.

    So apparently the provider updates the node and afterwards some VMs (a friends VM doesn't boot either) don't boot anymore, I guess that's totally the customers fault.

    Deja Vu.

  • irmirm Member

    MCHPhil said: It is well known it can take a while to do a node reboot. As I stated in the email sent before this happened. Starting tons of VM's at once takes time. It isn't a 10 minute process. Likely your VM was still booting.

    I understand that and I wasn't trying to ping/boot my VM as soon as the node came back online.

    I attempted to ping/access my VM 58 minutes after the second email was received (which is when I saw it) which was the email letting customers know that all nodes except "CA-VPS3" (which I am not on) were back online and that was when I ran into the issues I listed above.

  • tr1ckytr1cky Member
    edited April 2014

    @Virtovo said:
    Ask MCH to see if the image is being found when the domain starts. If using SolusVM they can do this quite easily using the debug boot option otherwise you can get the same info starting the domain from the command line.

    In the state I am right now, with vnc showing "No bootable device" I can't even type anything into VNC (and yes, I changed the drivers back and it didn't resolve anything).
    Meanwhile a mate get's this (vm also not booting since node restarts were performed):

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • MCHPhilMCHPhil Member, Provider

    @irm said:
    I attempted to ping/access my VM 58 minutes after the second email was received (which is when I saw it) which was the email letting customers know that all nodes except "CA-VPS3" (which I am not on) were back online and that was when I ran into the issues I listed above.

    There may have been some confusion with that email. That is my fault. It was meant to be taken as the host nodes are up and things are good, minus work being performed on CA-VPS3. I was recompiling some modules for the newer kernels. This required about 5-7 minutes of work per node. CA-VPS3 hung on shutdown (after disks were unmounted) or I just became inpatient and rebooted it via IPMI hence the longer time it took to bring it back online. The auto start of the VM's was also performed on the nodes at that time. I should have added a time frame to the second email. I thought roughly 30-45 minutes but was not confident enough to quote a timeframe.

    Thanked by 1irm
    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    @Virtovo said:
    Ask MCH to see if the image is being found when the domain starts. If using SolusVM they can do this quite easily using the debug boot option otherwise you can get the same info starting the domain from the command line.

    There are no errors on loading his VM. His image is found and active.

    lvdisplay /dev/vg_kvm/kvm671_img

    --- Logical volume ---

    LV Path /dev/vg_kvm/kvm671_img

    LV Name kvm671_img

    VG Name vg_kvm

    LV UUID lxcycm-4VN2-lOb2-Gl9a-U9kX-kJ12-G1xfs1

    LV Write Access read/write

    LV Creation host, time 2-99-16.net, 2014-03-28 20:18:16 -0400

    LV Status available

    # open 1

    LV Size 10.00 GiB

    Current LE 2560

    Segments 1

    Allocation inherit

    Read ahead sectors auto

    - currently set to 256

    Block device 253:122

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • @tr1cky said:

    @tr1cky said:

    I'd still ask for the host to do as I said before; however in the interim; have you and/or your friend tried booting a liveCD and inspecting the disk?

  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    @Virtovo said:
    I'd still ask for the host to do as I said before; however in the interim; have you and/or your friend tried booting a liveCD and inspecting the disk?

    image

    I suggested booting a recovery ISO, they are provided for these reasons.

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • @mchphil

    isn't it just easier to fix it for him instead of dealing with all those comments? either way it consumes your time.

  • Ian_Ian_ Member

    Yea boot the vps with a rescue disk with g parted to see if any of the partitions are showing up. If any of the partitions are showing up trying doing a re install of grub (assuming its linux) to restore your mbr. Sometimes not powering off the vm's manually before a reboot can cause problems like this. Yes i seen it happen to me a couple times where I didn't manually power off all the vps's before I rebooted one of my nodes.

    Tactical VPS | KVM and OpenVZ | FL (Go Rack) | L.A. (WebNX) | Dallas (Query Foundry) | https://billing.tacticalvps.net

  • tr1ckytr1cky Member
    edited April 2014

    @Ian_ said:
    Yea boot the vps with a rescue disk with g parted to see if any of the partitions are showing up. If any of the partitions are showing up trying doing a re install of grub (assuming its linux) to restore your mbr. Sometimes not powering off the vm's manually before a reboot can cause problems like this. Yes i seen it happen to me a couple times where I didn't manually power off all the vps's before I rebooted one of my nodes.

    Nope, no partition there.

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • NeoonNeoon Member

    My KVM is also still down and it was not the first time and he told me the same: I need to reboot it by myself because auto reboot failed.

  • @Infinity580 said:
    My KVM is also still down and it was not the first time and he told me the same: I need to reboot it by myself because auto reboot failed.

    And does it boot?

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • NeoonNeoon Member

    It does boot but network is broken.

  • Did someone already un-mount the disk image, and change the boot option from CD to Harddisk?

  • Yes, I did, while my vm still has no partition at all, a friends vm at mch seems to be pretty corrupted, fsck is showing thousands of errors.

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • VirtovoVirtovo Member
    edited April 2014

    @tr1cky said:
    Yes, I did, while my vm still has no partition at all, a friends vm at mch seems to be pretty corrupted, fsck is showing thousands of errors.

    Your friend's VPS sounds like it didn't shutdown before the node reboot.

  • @Virtovo said:

    Oh, and I guess that's a problem the client has to solve, right? Because it's totally caused by the client -.-'

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • NeoonNeoon Member
    edited April 2014

    @ErawanArifNugroho you cant boot from DVD? oh i didnt know that

    Sorry for my sarcasim but i am a bit pissed atm Yes Harddisk is first in boot order and the DVD is not mounted. Idk how the network can broke at reboots on KVM's and i got told reinstall it ....

  • VirtovoVirtovo Member
    edited April 2014

    tr1cky said: Oh, and I guess that's a problem the client has to solve, right? Because it's totally caused by the client -.-'

    I'm trying to help here.

  • I have VPS on Nodes CA2, CA3 and CA4. The three went down for reboot and the three came back up without any issues.

    Thanked by 1MCHPhil

    Affiliate link for one of my favorite provider's special offer page

  • InfinityInfinity Member, Provider

    @Infinity580 said:
    ErawanArifNugroho you cant boot from DVD? oh i didnt know that

    Sorry for my sarcasim but i am a bit pissed atm Yes Harddisk is first in boot order and the DVD is not mounted. Idk how the network can broke at reboots at KVM's and i got told reinstall it ....

    What do you mean by network broken? Can you give more details? Have you checked your network interface is set to be started on boot, some default installations don't have that option on?

    Cablestreet - London based ISP - Managed Solutions, Carrier Services, Colocation, Dedicated Servers, VMs, and more..

  • NeoonNeoon Member
    edited April 2014

    Let me explain that, after a reboot from client or node, eth0 dosent come up anymore. Thats what i know, VNC isnt working atm so i cant tell more. That happend 2 times in row, fresh reinstall didnt fixed it.

  • blackblack Member

    2 of my VMs on different nodes went down and came back up ok. Sorry to hear those that didn't.

    Thanked by 1MCHPhil
  • [Infinity580 said]eth0 dosent come up anymore. .. VNC isnt working atm so i cant tell more.

    I know what you feel :) We can't fix anything if we can't connect via vnc now. If you decided to do a new reinstall again, try to use static IP in the network, don't use DHCP.

    If the network still have problem, then DHCP is the last option.

  • I don't recommend using DHCP @ MyCustomHosting.

    Affiliate link for one of my favorite provider's special offer page

  • Ian_Ian_ Member

    @tr1cky I hope you had a backup! Sounds like a re installation is needed.

    Tactical VPS | KVM and OpenVZ | FL (Go Rack) | L.A. (WebNX) | Dallas (Query Foundry) | https://billing.tacticalvps.net

  • @Ian_ said:
    tr1cky I hope you had a backup! Sounds like a re installation is needed.

    Yeah, it's fine, I guess I won't hear an apology from @MCHPhil though.

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    @tr1cky, What is it I need to apologize for? I send a email out about the reboot. It's regular maintenance. You could have minimized the risk of this happening by shutting your VM down prior to the host node reboot. OR simply making sure nothing hangs when you restart your VM. I have 2-3 VM's PER node and not 1 had trouble rebooting. Some are running cPanel with tons of domains (300+), some are just simple DNS servers. EVERYTHING I run is on my infrastructure. That's how I'm so confident the maintenance I performed was not the cause of the problem. I or @black or @frankz or a multitude of others would also be having issues. But it is clearly that the VM's did not shut down in a timely fashion.

    Thanked by 10xdragon
    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • LukeTLukeT Member

    MCHPhil said: minimized the risk of this happening by shutting your VM down prior to the host node reboot

    You only gave like an hours notice?

    I should really get a better signature!
  • @MCHPhil said:
    tr1cky, What is it I need to apologize for? I send a email out about the reboot. It's regular maintenance. You could have minimized the risk of this happening by shutting your VM down prior to the host node reboot. OR simply making sure nothing hangs when you restart your VM.

    You send an email out 30 minutes before maintenance.

    And I quote:

    @Ian_ said:
    Sometimes not powering off the vm's manually before a reboot can cause problems like this. Yes i seen it happen to me a couple times where I didn't manually power off all the vps's before I rebooted one of my nodes.

    tsdns.io - free, redundant, DDoS-protected TSDNS

  • MCHPhil said: I send a email out about the reboot. You could have minimized the risk of this happening by shutting your VM down prior to the host node reboot.

    Is this a joke? You sent notification out few minutes before host node reboot.

  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    This is just cute, should have just left a few services running with a heartbleed vulnerability I guess? You could have even messaged me before hand and said hey could I have a little more time or etc. Could do that anywhere nowadays..

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • RocksterRockster Member
    edited April 2014

    So you heard about heartbleed vulnerability today and decided to reboot it immediately? That's the dumbest excuse I ever heard.

    You could have even messaged me before hand and said hey could I have a little more time or etc.

    Is this a joke again? You sent notification out few minutes before host node reboot.

  • LukeTLukeT Member

    Considering heartbleed has been known about for around 3 days, I'm sure more than 30 mins notice to a reboot was possible.

    I should really get a better signature!
  • MCHPhilMCHPhil Member, Provider

    LOL! Yeap rockster, today. :) Sometimes you guys are just great. Notice the sarcasm. Problem being one process from SolusVM was still sitting around. I had been playing with heartbleed against it and had it locked out via iptables.

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    @LukeT said:
    Considering heartbleed has been known about for around 3 days, I'm sure more than 30 mins notice to a reboot was possible.

    Possible, all things are. At this price point. It's not. On a saturday at that. Maintenance is pushed to weekend to keep uptime during the week. It's a reboot for christ sakes, MOST providers here don't even tell you who really owns them let alone when they are going to reboot a node. This is just a joke.

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • RocksterRockster Member
    edited April 2014

    This guy seems to think that 20 - 30 minutes of time in saturday is enough that every client see his mail and power down VPS...

  • MCHPhilMCHPhil Member, Provider
    edited April 2014

    @Rockster said:
    This guy seems to think that 20 - 30 minutes of time is enough that every client see his mail and power down VPS...

    You clearly want the world for $15 a year.

    I'm done here.

    Transparency is key - Webhosting with a spine! - Live US based support, not to be confused with aliens.
  • RocksterRockster Member
    edited April 2014

    MCHPhil said: You clearly want the world for $15 a year.

    So now I am the problem? Why you ALWAYS and I really mean ALWAYS blame others for your own incompetency. There's like gazillion reports at LET about issues caused by you, but you didn't say "I am sorry for mess I caused" even once.

This discussion has been closed.