The nodes with problems have been rebooted. Under observation.

NOC: To fix the disconnection issues, we still need to reboot all other nodes again. During 1400-1430 EST.

DMIT will set all 1vCore VM at LAX Datacenter to 2vCore. There will be no Single-Core VM in LA Datacenter for specific reasons.
If you have 1vCore, upgrade to more. You will get 1 more vCore.

RAM will also be increased to at least to 1GB, (no changes for the RAM larger than 1GB)
This will match the default setting, if you have 0.5G in the package, and 1G in extra, your configuration will be 2G.

Please note, it takes time to do this change. Please do NOT send a ticket, our system will do it automatically.

Also,
Please DISABLE swap in the system, first, we have IO Limited for each level products, if you used up all IOs and RAMs, the pages cannot be copied in or out which will lead to packet loss, performance drop and network disconnection.
If you run out your RAM, please upgrade your RAM. SWAP is not a solution in cloud platform.

All these changes will need you to do manually STOP and BOOT.

NOC DMIT

Since recently unstable Node servers, dozens of our clients faced unstable connections on the VMs. In order to show our determination to solve the problem

For the existing client who has activated order now, there is a life time 15off / 20off code we offered:

Semi-Annually(15off) or Annually(20off) and 1.3x Traffic / Bandiwdith in LAX.

Semi
Semi-Extend-Unstable-SLA-Existing-Client-LiftTime-Discount-For-Dec-2019-LAX-1.3x

Annually
Extend-Unstable-SLA-Existing-Client-LiftTime-Discount-For-Dec-2019-LAX-1.3x

New users 13off 2 times with 1.3x (Quarterly Semi-Annually Annually)
New_Users_LAX_DEC_1.3x

We got update from NOC, we already order some new stronger performace brand new Intel 2nd gen Scalable servers. All client in LAX.Pro will be moved to the new node servers.

This problem only exists in few numbers of VMs in our current 4CPU E5-4657Lv2 Nodes. Even if we do a lot of kernel optimization, this problem will still there. It has been alleviated but cannot be eliminated for the time being. So, we decided to replace all the servers in this series.


Sales.

HKG Scheduled Maintainance:

Since Insufficient resources and network upgrade, we are going to upgrade our HKG Core Network at Monday Dec 9, 2019( Hong Kong Time ). It takes hours, please be prepared.

Because the update of WHMCS conflicts with our API; This resulted in some VMs control panel connection errors and control panels were inaccessible. It has been fixed now.

PayPal Event:
Once you paid your invoice by PayPal, you will get a 6% cashback in DMIT Credit from today to Dec 20, 2019 for any LAX Invoice.

Sales.

At Dec 12 02:15AM, it takes hours.

Mostly done, except CW IPv6 Session, and CERNET IPv4 optimization.

Quick maintain for *HBG3W1 in HKG.

NOC will update the customize PVE-Firewall on all DMIT PVE nodes within 48 hours. The wrong design of PVE-Firewall leads to packet loss in the Linux Kernel(The ARP packet loss will lead to disconnection). We have fixed it in our test environment. After that, we will start to maintenance our own Proxmox branch under the stable master branch.

This progress is ongoing. It will not cause any interruption. The programs have been updated; the configuration of all VM will be updated in 90mins.

The disconnection issues have been fixed on both LAX and HKG
The new node still takes time for delivery, installation and testing.

Semi
Semi-Extend-Unstable-SLA-Existing-Client-LiftTime-Discount-For-Dec-2019-LAX-1.3x

Annually
Extend-Unstable-SLA-Existing-Client-LiftTime-Discount-For-Dec-2019-LAX-1.3x

New users 13off 2 times with 1.3x (Quarterly Semi-Annually Annually)
New_Users_LAX_DEC_1.3x

Anyone 20off 8 times with 1.3x (Annually)
LAX_DEC_20off_annual_1.3x

These promotions will end on Dec 29, 2019

For all users who cancel server renewal on LAX Datacenter due to instability during this period, we can assist you to restore the previous order at the original price.

The route of China Mobile AS9808 from China to Los Angles has been rerouted to Hong Kong.