Start a Conversation

Solved!

Go to Solution

Moderator

 • 

181 Posts

4410

December 23rd, 2021 14:00

STARTUP and SHUTDOWN - normal and what to do if it fails - all models

The goal of this discussion is to provide a place for everything related to STARTUP and SHUTDOWN for all models in one place.

There are a few useful discussions already which address parts of this, and I will include those links here as well.

This will be a work in progress and we welcome questions and contributions from all.

Lets get started...

 

SHUTDOWN:

BEFORE you shutdown:

*Make sure you either have physical access to the system or have a good iDRAC connection- so able to power-on again after shutdown. 

To check ahead of time for issues before initiating the shutdown:
1. Open a putty session to the Avamar node as 'admin' user:
2. Run these separately on AV as root and check output:
avosshutdown precheck
dpnctl shutdown --precheck






NORMAL SHUTDOWN:

The basic answer, whether a DP4400 or other IDPA model is this (order important):

  1. IMPORTANT- YOU MUST DO THIS FIRST:
    1. logon to the acm dashboard as root with the IDPA  common root pw (put name or IP or acm server in browser and add :8543 (i.e. https://IDPA_name_or_IP:8543 ) 
    2. Click the button in the upper right corner to select Shutdown Appliance (This will do a proper shutdown of each of the IDPA components, such as Avamar, Data Domain, etc so no data is lost or corrupted.)
  2. You should then use iDRAC for each esxi host to do a Graceful Shutdown, which will shutdown the esxi host(s) properly. (DP4400 has 1 esxi host, other models have 3 hosts).

Here are more detailed steps:


You MUST shut down the IDPA from the ACM console.
Prerequisites
● Ensure that there are no backup jobs running on the Avamar Backup Server.


If there are backup jobs running on Avamar Backup Server when the IDPA appliance shutdown operation is in progress, the shutdown operation waits for the Avamar jobs to complete with the status Waiting for shutdown of Backup Server. It is recommended that you wait for the backup jobs to complete. However, if you must shut down the appliance immediately, and then you must log in to the Avamar UI and cancel the backup jobs that are in progress.


● When shutting down IDPA, physical intervention or the use of iDRAC will be needed to restart the system. If you are remotely shutting down the IDPA appliances, ensure that either you have physical access to the system or have configured iDRAC on the system.


To shut down IDPA, perform the following actions:
Steps
1. On the ACM dashboard Home tab, click the Shutdown Appliance icon.
2. Enter the ACM root password, and click Yes.
The appliance shut down progress is displayed.
The IDPA appliance shuts down the components in the following order:
● Backup Server (Avamar)
● Search (Data Protection Search, DPS)
● Reporting and Analytics (Data Protection Advisor, DPA)
● IDPA System Manager (Data Protection Centeral, DPC)
● Cloud Disaster Recovery Agent
● Protection Storage (Data Domain)
● Appliance Configuration Manager (ACM)
● vCenter Server (VCSA)
● Compute Node (ESXI)














If the ACM-initiated shutdown fails to shut down the Backup Server and the Protection Storage or both, then the ACM displays a message listing the component(s) that failed to shut down. The ACM then continues to shut down the other components.


For DP4400, after all the point products (VMs) shut down, the ESXi needs 15 minutes to go into maintenance mode and shutdown.
If the ESXi does not shut down, you need to manually shut down the ESXi. DP4400 takes more than 30 minutes to shut down. For more information about shutting down ESXi manually , see Shut down ESX manually in the Product Guide.
If any components fail to shut down, you must manually shut down the components. For more information about manual shut down of IDPA components, see Troubleshooting shut down in the Product Guide and/or see below.


IF NORMAL SHUTDOWN FAILS:

Troubleshoot shutdown


During the shut down process, if the appliance or any of the components fail to shut down automatically, you can manually shut down the IDPA appliance and its individual components.
If the Avamar or Data Domain systems fail to shut down, aside from performing a manual shut down of these components, you must also shut down the infrastructure components, which include vCenter and ESX servers.

 

If the shutdown is already hung, you can putty to the acm server to check the logs:

/usr/local/dataprotection/var/configmgr/server_data/logs/server.log
You could grep for "Avamar shutdown precheck failed"


Avamar shutdown validation errors and suggested actions:

DELLRodK_0-1640297099271.png

DELLRodK_2-1640297172393.png


Shut down Avamar manually

This section provides you information about how to shut down the Avamar component manually in the DP5300, DP5800, and DP4400 appliances.


To manually shut down the physical Avamar (DP8xxxx only) or Avamar Virtual Edition (AVE) (DP4400/DP5xxxx) component, perform the following actions.

Login to the Avamar or  AVE server with SSH by using the Avamar/AVE IP address.
a. Create a checkpoint by running the following command.
mccli checkpoint create --override_maintenance_scheduler
b. Backup the MCS data by running the following command:
mcserver.sh --flush
c. Stop all Avamar services by running the following command.
dpnctl stop all
d. Power off the physical Avamar or if the AVE then power off from vCenter.
NOTE: If you have the DP4400 or DP5xxxx appliance, then you must shutdown ESX(s) if full HW power down is desired. For more information about manual shut down of ESX, see Shut down ESX manually in the Product Guide and/or see below.














Shut down Data Domain manually


This section provides you information about how to shut down the Data Domain component manually.

To manually shut down the Data Domain component or the DDVE, perform the following actions.
NOTE: The manual shut down for Data Domain system is applicable for the DP5xxxx and DPxxxx appliances.
NOTE: The manual shut down for DDVE is applicable for the DP4400 appliance.
Steps
1. Open a new SSH session to login to the Data Domain system or DDVE.
2. Shut down the Data Domain or DDVE system by running the following command.
system poweroff
NOTE: If you have the DP4400 appliance, then you must shutdown ESX if full power-down is desired. For more information about manual shut down of ESX, see Shut down ESX manually in the Product Guide, or see below.










Shut down VMs in vCenter manually:


This procedure is applicable for all models of IDPA. Perform the following actions:

1. Open a browser and enter the IP address to access vCenter.
The login page is displayed.
2. Enter your username and password on the vCenter login page.
3. Shut down the Data Protection virtual application (vApp).
NOTE: All virtual machines and virtual applications under Data Protection virtual application are automatically shut down in the proper order. (does NOT include vcsa and acm)
4. If needed after above then the VCSA can be shutdown by connecting to the esxi host client for the esxi hosting the VCSA VM, and Shutdown the VM.
5. If needed after above then the ACM can be shutdown by connecting to the esxi host client for the esxi hosting the ACM VM, and Shutdown the VM.








 

Shut down ESXI manually

This procedure to shut down ESX manually is applicable for all models of IDPA. To manually shut down the ESX, perform the following actions.

1. Login to each ESXi host.
2. Set all ESXi hosts into maintenance mode by running the following command on each host at the host console:
esxcli system maintenanceMode set -e true -m noAction
3. Shut down all ESXi host using the vSphere client or the ESXi host client.
4. If issues with above you can connect to iDRAC for each ESXI host and select Graceful Shutdown (NOT Power Cycle or Reset etc) on the dashboard.
(Only do this after proper shutdown from ACM dashboard and other steps above have been tried first.)






 

In the next post we will look at STARTUP......

 

We welcome questions and contributions, and I will post more....

 

If you found this useful please give KUDOS to help others find the information also.

thanks

Moderator

 • 

181 Posts

January 20th, 2022 15:00

Replying to mark TOTAL OF ALL PREVIOUS POSTS IN THIS DISCUSSION as the SOLUTION.

But please continue to post any questions/comments.

Moderator

 • 

181 Posts

December 31st, 2021 10:00

Continued from my previous post ......(for SHUTDOWDN see my previous post)

Now lets look at STARTUP...

 

STARTUP:

 

NORMAL STARTUP:

DP4400 only:

  • For normal startup, switch on the power button present on the Dell Server (DP4400). The ESXi internally starts local.sh (/etc/init.d/local.sh) and the VMs should start automatically.
  • Allow at least 2hrs 30mins for complete functionality.
  • After PowerProtect starts successfully, you can connect to the ACM dashboard and monitor the progress of the startup. If there is a failure in the startup, the application displays an error message with an option to access the ACM dashboard page. 
  • If issues, see Startup Troubleshooting steps below.

 

DP5xxx/8xxxx only:  

Powering on PowerProtect requires starting individual components in the *correct order.  

1. Power on Data Domain system.  

 

2. Power on the NDMP accelerator or accelerators if they are included in the configuration.

  • **Do not continue until both Data Domain and all NDMP accelerators are fully initialized.
  • To verify that the Data Domain system is fully initialized, log in to the Data Domain system using sysadmin as the username , and for password, use the PowerProtect common password. Check the file system status using the filesys status command. The file system should be up and running.
  • To verify that the NDMP accelerators are fully initialized, connect to the accelerator node using SSH.  

 

3. For configurations with a physical Avamar implementation, power on the utility node and storage nodes. This step is not necessary for configurations with AVE.  

  • Login to the Avamar Admin GUI to ensure that the appliance is powered on and Avamar is fully initialized.  

 

4. Power on each of the ESXi servers

  • Moving from the bottom ESXi server to the top
  • Once the ESXi servers are powered on, PowerProtect automatically starts all other components (vApps/VMs).  

 

5. After PowerProtect starts successfully, you can connect to the ACM dashboard and monitor the progress of the startup

  • Allow at least 2hrs 30mins for complete functionality.
  • If there is a failure in the startup, the application displays an error message with an option to access the ACM dashboard page. 
  • If issues, see Startup Troubleshooting steps below.

 

 

IF NORMAL STARTUP FAILS:

Troubleshooting startup


If one part of the startup process fails to complete automatically, the problem can be resolved manually to enable startup to continue.


Verify all physical nodes are powered on and have network connectivity

(If needed, use iDRAC to power on, verify any HW issues and fix those first if- depending on the issue- it may prevent full startup and operation. If a redundant HW component fails, then that can be addressed later.)


If the VMs do not start- start the VMs manually in order as follows:

(Below mentions AVE and DDVE but these are only included in certain models- otherwise they are physical nodes.)

1. Log in to ESXi using idpauser or root and select Exit maintenance mode.

 

2. Start the DataProtection-VCSA VM by running the /etc/init.d/local.sh script on ESXi or power on the VM from the ESXi.

 

3. DataProtection-ACM VM should start five minutes after the VCSA VM starts. If it does not start automatically after waiting long enough, then power it on from the ESXi or vCenter.

 

4. If DDVE VM is not started, power it on from ESXi or vCenter.
Wait until DDVE VM is up and running before next steps.

 

5. If AVE is not started, power it on from ESXi or vCenter.

  • Verify Services are running as follows:
    • Log in to AVE using admin credentials.
      Run dpnctl status all and based on results- if needed run dpnctl start all, and dpnctl start maint commands. (Wait until all services are up before doing next steps.)

 

6. Power-on the below *in order from ESXi or vCenter:

● DataProtectionSearch Vapp
● DPADatastoreServer VM
● DPAApplicationServer VM
● DataDomainCloudDR VApp
● DataProtectionCentral VApp
● AVProxy VM

 

7. If DPS services do not start:

Log in to Index Primary IP using operating system root credentials and run the following commands:

  • service elasticsearch start
  • service search-cis-core start
  • service search-cis-schedule start
  • service search-networker-worker start
  • service search-networker-action start
  • service search-avamar-worker start
  • service search-avamar-action start
  • service search-worker start
  • service search-adminapi start
  • service search-api start

 

8. If DPA services do not start:

  • Log in to the DPA Application Server VM using SSH.
  • Stop all DPA Application Services by running the following command:
    /opt/emc/dpa/services/bin/dpa.sh service stop
    If there are any errors, you may continue to ignore them.
  • Verify by running: /opt/emc/dpa/services/bin/dpa.sh service status
  • Log in to the DPA Datastore Server VM using SSH.
  • Stop all DPA Datastore Server services by running the following command:
    /opt/emc/dpa/services/bin/dpa.sh service stop
  • Verify by running: /opt/emc/dpa/services/bin/dpa.sh service status
  • Once all the services on both the DPA Application and the DPA Datastore servers are stopped, restart the services on the DPA Datastore Server by running the following command:
    /opt/emc/dpa/services/bin/dpa.sh service start
  • You must wait for some time until all the services are started on DPA Datastore Server. You can verify the status by running the following command:
    /opt/emc/dpa/services/bin/dpa.sh service start.
  • Once all the services on the DPA Datastore server are started, restart the services on the DPA Application Server by running the following command:
    /opt/emc/dpa/services/bin/dpa.sh service start
  • You must wait for some time until all the services are started on DPA Application Server. You can verify the status by running the following command:
    /opt/emc/dpa/services/bin/dpa.sh service status

 

9.  Power on the AVProxy VM from ESXi or vCenter.


10. After all above started, then start tomcat (web) services of Avamar using dpnctl start emt (if not already running).

 

That's it for Startup.

If still having issues you can search for answers or open a Dell Support case here:

https://www.dell.com/support/home/en-us

 

 

We welcome questions and contributions, and I will post more....

 

If you found this useful please give KUDOS to help others find the information also.

thanks

1 Message

June 21st, 2023 09:00

Hello Rod,

thanks for your procedure, it is very detailed.

is there any other way to remote shutdown the system rather using ACM UI or individual component shutdown? I guess the button on ACM UI is calling a script on the ACM so, can customer can call it remotely if so?

Moderator

 • 

181 Posts

June 28th, 2023 08:00

That is a great question @MarioCavero .

I have never seen or heard of that option so I don't think the assumed script or executable is officially published,  so I will assume that is not an option unless I can find out otherwise.

But I am looking further into it and will reply if I find anything on that.

Otherwise, in-case acm GUI is not available to do the proper shutdown, these steps can be done to try to get the GUI up:

  • To restart the web GUI on ACM:
    service dataprotection_webapp status
    service dataprotection_webapp stop
    service dataprotection_webapp start
  • Other options to check it:
    systemctl status dataprotection_webapp
    netstat -anp | grep -w 8543
    ps -ef |grep apache

I hope this helps.

 

We welcome questions and contributions from all.

If you found this useful please give KUDOS to help others find the information also.

thanks

No Events found!

Top