Node 7SYJ422 Reboot for kernel update

Dear Customers,

These are a few very important notices. Please be sure to read them.

1. We will upgrade the architecture of Hong Kong Data Center from December 29 to January 1, Hong Kong time. Please make sure to backup your data. ( all customers )
- In the meantime, we will upgrade each node one by one. Because we need to handle data security. Cannot batch process.
- At the same time, you will unable to manage your server through a web page.
- We can ensure that your service is less affected by this upgrade, because your server will not be available until the node where your server is located is upgraded.

During this period, your node server will be offline for about 3 hours, and we will choose to do it in the morning or early hours of Hong Kong time to ensure that your business will be less affected.


2. We will upgrade the architecture of each node server as follows:

The server will be redeployed during this period, so be sure to back up your data.

The user data of PVM.n2 will be saved, but there is a risk of data loss.

The user data of PVM.n3 cannot be saved. As PVM.n3 is an Open Beta product, please understand.

- Each node server will be upgraded to 2Gbps LACP;
- Upgrade to PVE architecture;
- Support for custom disks, IP numbers, etc.;
- Support VM automatic backup ( snapshot );
- Support Cloud ACL;
- Optional networks ( N2 | standard, n3 | Premium, IPv6-only, Global ), but not both;

3. Adjustment of Cart and Product structure:
- First of all, there will be no change in your package content, renewal price, discount, etc.
- From January 2019, we will use the following method to divide products:
A ) Burst Instances ( Xeon E5, IVY, DDR 3 )
B ) Common Instances ( Xeon E5, IVY, DDR 3 )
C) RAM instances ( Xeon E5, haswell, DDR 4 )
D ) Dedicated instances ( Xeon E3, Sky Lake / Kaby Lake, DDR 4 ). Each CPU Core is non-shard allocated to vCore.

4. After the architecture upgrade is completed, all existing VM's ( PVM.n2 and PVM.n3 ) will use the B ) common instance model by default.

5. PVM.n3 package will be adjusted;
- The configuration will be the same as the current PVM.n2 configuration ( including network, core and disk, etc. ); Prices will change you selects N3 Premium network in cart.
- Existing PVM.n3 users can upgrade to the new package or keep it.

6. PVM. N2 / N3. Large package will be offline and the existing package will be renewable

7. Hong Kong's network will be upgraded from March 10 to 15, 2019. Including N2 Standard and N3 Premium. We will connect AS 2914 NTT US.

Best regards,

DMIT INC MANAGEMENT AND NOC DEPT.

Please save your data; In the process of upgrading, there are still unforeseen circumstances leading to data loss.

NTT will become one of our upstream ISP. PCCW/ANCHNET still will be the main ISP we connect to China.

Dear customer,

Please note that we are going to start the upgrade process of Hong Kong A Datacenter.

We will start at PVM.n3 VM. Lastest backup time: 18:00 Dec 30 2018. The network of PVM.n3 will be disconnected. you will also unable to manage your server through our web or SSH.

Please, backup all your data, including PVM.n2 user. After we finished the upgrade of PVM.n3, we will start the upgrade of PVM.n2.

During this time, we will not deliver the new PVM order. Thanks

Best regards,
DMIT NOC

Dear Customers,

According report,
IPv6 Only Products will not upgrade to PVE this time.

The IPv6 network will temporarily unavailable after upgrade to PVE. Because we have to allocate to every VM /64 subnets, but PVE and PVE modules doesn’t support it now. We are going to modify it.

Best regards,
NOC

Dear all,

There is an unknown issue on the Node Switch. We will go to check in Datacenter within 3 days.

- Due to the Node Switch of PVM.n2 out of management. We can only go to the Datacenter to re-configure the Node Switch.

- At the same time, we will replace the Core-Switch to Cisco Nexus 3000.


- We also start the migration for ALL VM. At now, you cannot place the order to have new VM. But you can use your VM normally. The migration will take about 10-30 mins(depends on your VM disk). It will only take effact to your VM when your VM migrating.

From Gateway to Looking Glass in N2 Server

This is due to STP settings reverting to default, an then resulting in STP conflicts(Resulting in constant reconvergence of STP network topology.)

PVM.N2 Packet Lost & High Unknow RX Traffic has been fixed. Please check it again.

Dear All,

HK Time Jan 12 2019 2AM - 2:30 AM, we are going to switch the [Core Switch] to [Cisco N3k], there will be 30 mins downtime.

We are sorry for the unstable network during this time.

Please understand that we have to upgrade our network one step by one step.

Best regards,
DMIT NOC.

Maintenance report:

All of our expected goals have been achieved, and we will adjust the network architecture of N2 and N3.

Best,
NOC

Dear Customers,

We will terminate HKT's service one month in advance; If you are a new user who has ordered for less than 7 days, you can apply for a PAYPAL refund. if you have ordered for more than 7 days ( Include ) and renewed, the remaining days will be returned to your DMIT Credit by proportion.

Best regards,
DMIT SALE DEPT.

For reference

The BGP session of the core router failed unexpectedly. We are contacting upstream to repair.

Fixed

We have dozens of nodes waiting for migration.

After the migrated completely, your traffic will be reset. You can use it normally.

We only use GE port for migration. According to the average rate ( 90MB/s ), a server with 10G hard disk needs 115 seconds to copy the files completely. At the same time, it takes 30 seconds for API operation.

During this period, please !!!DO NOT!! go to the console to BOOT your server. We have a log record. If you BOOT the server during migration and make your data loss, we have no responsibility for that.

BEST,
NOC

PLEASE DO !N!O!T! BOOT SERVER EVEN YOUR CANNOT ACCESS SERVER. PLEASE SEND TICKET ABOUT THAT

Dear customer,

Since now, all of PVM.n2 were been migrated to PVE architecture.

From now on, we will push forward the solution to the new platform problem step by step. After these we will send your a completely email content.

All you wanna know is that:

All the VM which migrated from SVM platform does not support to do following actions.
- Reset root password via control panel;
- Soft shutdown;
- Automatically setting. (IP, SSHKEY, HOSTNAME, ROOT PASS)
- accurate CPU/RAM usage data.

All of these gonna be fix after you follow our guide.

If your VM migrated from the previous platform and you forgot your root password, at the same time you don't want to reinstall the system. You can follow this guide to reset your password:

https://www.dmit.io/index.php/knowledgebase/6/---cloud-init.html