Version upgrades

Version upgrades#

Replicator will continuously be updated to add new features, enhance existing functionality, and/or apply patches for security threats. All update packages will have been applied to Plixer’s QA servers and thoroughly tested before being made available to the public.

Note

Replicator update packages can be installed without assistance, but it is strongly recommended to contact Plixer Technical Support to allow our support engineers to guide you through the process.

On this page:

Pre-upgrade checks/preparations

Replicator 20.1.1

HA pair upgrades (from v19.1.1)

Upgrading v19.1.1 HA pairs

Pre-v19.1.1 upgrades

Upgrading to v19.1.1

Pre-upgrade checks/preparations#

The checks below are meant to identify and resolve any underlying issues with the current Replicator install and should be completed before the scheduled upgrade window to help ensure that the upgrade will be successfully completed.

All install logs are saved to /var/log/Replicator-Install.log.

All upgrades#

[ ] A VM snapshot of the standalone/configuration instance (v20.0.0+) or the appliance to be upgraded (pre-v20.0.0 installs) is available.
[ ] Web interface and SSH (replicator user for v19.1.1 or plixer user for v20.0.0+) credentials are available for each Replicator instance to be upgraded.
[ ] Vitals LEDs in the web interface Admin section are green.
[ ] The root disk has at least 5 GB of free space:
```
df -h /
```
[ ] The upgrade path and estimated duration have been confirmed against the upgrade path table.

Offline upgrades#

[ ] The offline packages have been downloaded, moved to /var/db/big/offline/ (requires an additional 84 GB), and checksum-verified. Allow extra time to download/transfer the offline bundle due to its size. If necessary, add more storage or use/deploy a Scrutinizer server as a temporary offline repository host.
[ ] If using an active Scrutinizer primary reporter for the offline repository, 10% of its total disk space must remain available after the offline packages are downloaded and extracted. This will prevent automatic history trimming. The offline packages should not be staged on the root filesystem as they can fill the root disk and stop flow collection.

Situational checks#

Each check below leads with the situation it applies to. Skim the bold labels and complete only the checks that match the deployment being upgraded.

[ ] Version upgrades - Verify that the Replicator appliance meets the target version’s minimum resource requirements.

Important

Do not proceed with the upgrade if the server is underprovisioned. Contact Plixer Technical Support for assistance with tuning system parameters or allocating additional resources to the server.
[ ] Headless conversion (upgrades from v19.1.1 only) - Log in to the configuration instance and register/add a new headless instance using the IP address that will be assigned to the v19.1.1 appliance after it is upgraded and converted to a headless instance.
- [ ] v19.1.1 configuration migration - Back up the v19.1.1 configuration data (to be restored on the v20.0.0+ configuration instance after the upgrade and conversion to headless)
[ ] Internal NTP server - If the server being upgraded to Replicator 20.0.0+ cannot access the Internet, an internal NTP server can be configured by running the following:
```
sed -i -e '/^pool/aserver NTP_ADDRESS' -e 's/^pool/#&/' /etc/chrony.conf
```
[ ] Online upgrades - Confirm that the Replicator appliances to be upgraded can reach https://files.plixer.com by downloading the checksum file:
```
curl -O https://files.plixer.com/plixer-repo/scrutinizer/19.8.1/replicator-install.run.sha256
```
For deployments without Internet access, download the file from the IP address (REPO_HOST_IP) of the offline yum/dnf repository host instead.
[ ] High-availability pairs - Stop the keepalived service on each node before upgrading:
```
sudo systemctl stop keepalived
```
When upgrading v20.0.x HA pairs, the secondary instance must be upgraded before the primary. V20.1.x pairs can be upgraded in any order.
[ ] Redeploying from older versions - Obtain a valid license key for the target version if one has not already been acquired.
[ ] Previously upgraded servers - Delete any older versions of replicator-install.run on the Replicator instance. This prevents them from being used instead of the latest/correct installer.
[ ] Hardware appliances - Log in to iDRAC and perform a hardware health check. Escalate any issues found to Dell for resolution. A reboot is also recommended as an additional check for underlying hardware issues.

Once all applicable checks are complete, follow the instructions in the next section to upgrade Replicator to the latest version.

Changed in version 20.0.0: plixer is the recommended OS user for command line access. The replicator and root users are no longer required (the replicator user will be retained on instances that were upgraded from v19.1.1).

Replicator 20.1.1#

The standard procedure upgrades a compatible Replicator instance (see below) to the latest release.

Important

Replicator 19.1.1 appliances that are upgraded to v20.0.0+ default to standalone/configuration instance mode.

If an appliance needs to be converted to a headless instance, refer to these instructions before starting the upgrade process.

Upgrade paths#

The Replicator 20.1.1 upgrade requires the target VM or hardware appliance to be running v19.1.1 or higher and have at least 2 CPU cores, 8 GB of RAM, and 16 GB free on /.

Instances on older versions must be upgraded in stages, as indicated below.

Starting version	Upgrade path	Estimated time
20.0.x	20.1.1	30 - 45 minutes
19.1.1	20.1.1	60 - 90 minutes
19.1.0	19.1.1 -> 20.1.1	~1 hour
19.0.x	OS update + 19.1.1 -> 20.1.1	~3 hours
AWS 19.0.x	Must be redeployed as a 20.1.1 instance	30 minutes + setup and configuration
18.x and older	Must be redeployed as a 20.1.1 instance	30 minutes + setup and configuration

All upgrades from Replicator 19.1.1+ to the latest version are installed following the same procedure. For upgrades from older versions to v19.1.1, follow the links in the table above.

Backing up configuration data (upgrade from v19.1.1 -> convert to headless only)#

If a v19.1.1 instance is going to be converted to headless mode post-upgrade, its existing configuration data (profiles, collectors, and exporters) can be backed up before the upgrade. The backup data can then be restored on the v20.0.0+ configuration instance after the upgrade and conversion have been completed.

Installing the upgrade#

Once all preparation steps have been completed, the Replicator 20.1.1 installer can be downloaded and installed as follows:

Offline upgrades#

Replicator servers that cannot access the default yum/dnf repository on https://files.plixer.com can be upgraded by setting up an offline repository on the local network. This repository can be hosted on the Replicator standalone/configuration instance or a Plixer One/Scrutinizer server.

Warning

When using an active Scrutinizer primary reporter as the offline repository host, always download and stage the offline bundle under /var/db/big/offline/. Downloading it to the root filesystem (e.g., to /home/plixer) can fill the root disk and stop flow collection.

Before starting, verify that /var/db/big/ has at least 84 GB of free disk space:

df -h --output='avail' /var/db/big/

If necessary, add more storage or deploy a temporary Scrutinizer server to host the offline repository.

Upgrading v19.1.1 HA pairs#

To upgrade a Replicator 19.1.1 high-availability (HA) pair to v20.1.1, follow these steps:

Note

To set up a new v20.0.0+ HA pair, follow this guide instead.

Click to view instructions

Important

If the HA pair to be upgraded is using a virtual IP address (VIP), note the address being used (can be found in /etc/keepalived/keepalived.conf or by running ip addr show) before starting the upgrade process. The VIP will be required to complete several of the steps below.

Stop and disable the keepalived service.

sudo systemctl stop keepalived
sudo systemctl disable keepalived

Verify that the VIP was released by the primary and has been taken over by the secondary instance (run on both):
```
ip a
```
Upgrade the primary instance following this guide.
After the upgrade, log into the web interface and verify that profiles were successfully migrated to the upgraded primary instance.
Apply a new license (must support at least two Replicator instances).
Register a new instance under Admin > Resources > Replicators and note the API authentication token.
Reboot the primary instance, and then upgrade the secondary instance following the same steps as before.
After the upgrade, run the following on the secondary instance to set it up as a headless instance:
```
MAKE_HEADLESS=1 /usr/share/replicator/util/setup.sh
```
Enter the following details when prompted:
- Configuration instance (primary instance) IP address
- API authentication token created when the new instance was created (step 6)
- Name assigned to the new instance
Restart the secondary instance, and then return to Admin > Resources > Replicators in the web interface to verify that the secondary instance has successfully registered itself with the primary (IP address should be displayed).
On the same page, click on the primary instance name (usually ‘Local’), and configure high availability using the primary and secondary instances (enter the original VIP if necessary).
Click the Save button, and then wait two minutes.

Verify that /etc/keepalived/keepalived_replicator.conf contains the following (on both devices):

# this file is automatically generated do not edit
global_defs {
        script_user plixer
        enable_script_security
}
vrrp_instance VI_1 {
        state BACKUP
        interface eth0
        virtual_router_id 51
        priority 90
        advert_int 1
        preempt_delay 5
        notify  /usr/share/replicator/bin/notify.sh
        authentication {
                auth_type PASS
                auth_pass zhO2brnVF2byl4+zL/ssTaCPmHFr5IMQJEIVDkJiFCCvBIttqjBptwm8c8PWbjqqN4BVieKunkVqPWIRp4fG0Q==
        }
        virtual_ipaddress {
                10.42.150.17
        }
}

Modify /etc/keepalived/keepalived.conf on both instances:

sudo bash -c 'echo "include /etc/keepalived/keepalived_replicator.conf" > /etc/keepalived/keepalived.conf' 

Restart the keepalived service on both instances:
```
sudo systemctl restart keepalived
```

Both Replicator instances should now be running v20.1.1 and set up as a high-availability pair using the same (pre-upgrade) configuration.

Upgrading to v19.1.1#

The Replicator v19.1.1 upgrade requires a minimum of v19.0.1 due to an OS update:

V19.0.1 appliances - Update the OS before installing the upgrade as described in the guide below
V19.1.0 appliances - Skip the OS update and proceed to installing the v19.1.1 upgrade.

For assistance or clarifications, contact Plixer Technical Support.

Pre-upgrade preparation#

[Hardware appliances] Create a backup of the current Replicator appliance using the backup interactive mode command (will be saved in /home/replicator/backups) and store it on an external system/drive.
[Virtual appliances] Backup the current Replicator install by taking a VM snapshot.
Confirm the current password for the replicator SSH user (run passwd replicator)

Verify that root login is disabled by running:

sudo sed -i 's/^#PermitRootLogin.*/PermitRootLogin no/g' /etc/ssh/sshd_config

[Offline upgrades] If the Replicator appliance does not have access to files.plixer.com, set up an offline repository for this upgrade.
[High-availability pair] Stop the keepalived service before upgrading each instance (primary before secondary):
```
sudo systemctl stop keepalived
```

Upgrade process#

The process of upgrading a v19.0.1 Replicator appliance to v19.1.1 involves the following steps:

Backing up the v19.0.1 database and server-specific files using the backup interactive mode command.
Downloading the operating system upgrade script, olmigrate.run, and running it a total of four times (with a reboot between runs)
Downloading and running the Replicator v19.1.1 installation script (replicator-install.run)
Verifying that the v19.0.1 data has been successfully migrated after v19.1.1 is installed

Important notes#

The upgrade will take at least one hour to complete.
If the Replicator appliance is able to access files.plixer.com, the REPO_HOST variable should be set to files.plixer.com for the steps outlined below. For offline upgrades, the IP address of the offline repo should be used instead.
Due to an increase in minimum specs, older VMs with 2 GB of RAM should be provisioned with 4 GB RAM before the upgrade.
The upgrade requires a minimum of 16 GB free space on root (/). There may be older logs (sudo rm /var/log/messages-*) that can be deleted to free up space.

Updating the OS#

Once all preparation steps have been completed, follow these steps to update the OS to Oracle Linux 9:

Offline upgrades to v19.1.1#

The following instructions for setting up an offline repo are intended for upgrading to Replicator v19.1.1 only:

Version upgrades

Contents

Version upgrades#

Pre-upgrade checks/preparations#

All upgrades#

Offline upgrades#

Situational checks#

Replicator 20.1.1#

Upgrade paths#

Backing up configuration data (upgrade from v19.1.1 -> convert to headless only)#

Installing the upgrade#

Offline upgrades#

Upgrading v19.1.1 HA pairs#

Upgrading to v19.1.1#

Pre-upgrade preparation#

Upgrade process#

Important notes#

Updating the OS#

Offline upgrades to v19.1.1#