Upgrading InfluxDB Enterprise clusters
Upgrading InfluxDB Enterprise 1.3.x-1.5.x clusters to 1.6.6 (rolling upgrade)
Step 0: Back up your cluster before upgrading to version 1.6.6.
Create a full backup of your InfluxDB Enterprise cluster before performing an upgrade. If you have incremental backups created as part of your standard operating procedures, make sure to trigger a final incremental backup before proceeding with the upgrade.
Note: For information on performing a final incremental backup or a full backup, see the InfluxDB Enterprise Backup and restore documentation.
Upgrading meta nodes
Follow these steps to upgrade all meta nodes in your InfluxDB Enterprise cluster. Ensure that the meta cluster is healthy before proceeding to the data nodes.
Step 1: Download the 1.6.6 meta node package.
Meta node package download
Ubuntu & Debian (64-bit)
wget https://dl.influxdata.com/enterprise/releases/influxdb-meta_1.6.6-c1.6.6_amd64.deb
RedHat & CentOS (64-bit)
wget https://dl.influxdata.com/enterprise/releases/influxdb-meta-1.6.6_c1.6.6.x86_64.rpm
Step 2: Install the 1.6.6 meta nodes package.
Meta node package install
Ubuntu & Debian (64-bit)
sudo dpkg -i influxdb-meta_1.6.6-c1.6.6_amd64.deb
RedHat & CentOS (64-bit)
sudo yum localinstall influxdb-meta-1.6.6_c1.6.6.x86_64.rpm
Step 3: Restart the influxdb-meta
service.
Meta node restart
sysvinit systems
service influxdb-meta restart
systemd systems
sudo systemctl restart influxdb-meta
Step 4: Confirm the upgrade.
After performing the upgrade on ALL meta nodes, check your node version numbers using the
influxd-ctl show
command.
The influxd-ctl
utility is available on all meta nodes.
~# influxd-ctl show
Data Nodes
==========
ID TCP Address Version
4 rk-upgrading-01:8088 1.5.x_c1.5.y
5 rk-upgrading-02:8088 1.5.x_c1.5.y
6 rk-upgrading-03:8088 1.5.x_c1.5.y
Meta Nodes
==========
TCP Address Version
rk-upgrading-01:8091 1.6.6_c1.6.6 # 1.6.6_c1.6.6 = 👍
rk-upgrading-02:8091 1.6.6_c1.6.6
rk-upgrading-03:8091 1.6.6_c1.6.6
Upgrading data nodes
Repeat the following steps for each data node in your InfluxDB Enterprise cluster.
Step 1: Download the 1.6.6 data node package.
Data node package download
Ubuntu & Debian (64-bit)
wget https://dl.influxdata.com/enterprise/releases/influxdb-data_1.6.6-c1.6.6_amd64.deb
RedHat & CentOS (64-bit)
wget https://dl.influxdata.com/enterprise/releases/influxdb-data-1.6.6_c1.6.6.x86_64.rpm
Step 2: Stop traffic to the data node
Your cluster’s load balancer distributes read and write requests among data nodes in the cluster. If you have access to the load balancer configuration, stop routing read/write requests to the data node server (port 8086) via your load balancer before performing the remaining steps.
If you cannot access the load balancer configuration, we recommend working with your networking team to prevent traffic to the data node server before continuing to upgrade.
Step 3: Install the 1.6.6 data node packages.
Data node package install
When you run the install command, your terminal asks if you want to keep your current configuration file or overwrite your current configuration file with the file for version 1.6.6.
Keep your current configuration file by entering N
or O
.
The configuration file will be updated with the necessary changes for version 1.6.6 in the next step.
Ubuntu & Debian (64-bit)
sudo dpkg -i influxdb-data_1.6.6-c1.6.6_amd64.deb
RedHat & CentOS (64-bit)
sudo yum localinstall influxdb-data-1.6.6_c1.6.6.x86_64.rpm
Step 4: Update the data node configuration file.
The first official Time Series Index (TSI) was released with InfluxDB v1.5. Although you can install without enabling TSI, you are encouraged to begin leveraging the advantages the TSI indexing offers.
Add:
- If enabling TSI: index-version = “tsi1” to the
[data]
section. - If not enabling TSI: index-version = “inmem” to the
[data]
section.- Use ’tsi1’ for the Time Series Index (TSI); set the value to
inmem
to use the TSM in-memory index.
- Use ’tsi1’ for the Time Series Index (TSI); set the value to
- wal-fsync-delay = “0s” to the
[data]
section - max-concurrent-compactions = 0 to the
[data]
section - pool-max-idle-streams = 100 to the
[cluster]
section - pool-max-idle-time = “1m0s” to the
[cluster]
section - the [anti-entropy] section:
[anti-entropy]
enabled = true
check-interval = "30s"
max-fetch = 10
The anti-entropy service can initially be CPU intensive if you have very large shards (>50GB) and/or a large number of shards (>500) associated with your database. This can result in the cluster appearing to be unresponsive upon startup while shard digests are calculated. Once the shard digests are calculated, the overall CPU utilization should return to a lower level.
To check the number of shards within your cluster, use the
SHOW SHARDS command.
Use your operating system tools to check the shard sizes.
You may wish to test this configuration setting in a non-production environment to observe the system resource requirements prior to activating this service in production.
Remove:
max-remote-write-connections
from the[cluster]
section[admin]
section
Update:
- cache-max-memory-size to
1073741824
in the[data]
section
The new configuration options are set to the default settings.
Step 5: [For TSI Preview instances only] Prepare your node to support Time Series Index (TSI).
- Delete all existing TSM-based shard
index
directories.
- Remove the existing
index
directories to ensure there are no incompatible index files. - By default, the
index
directories are located at/<shard_ID>/index
(e.g.,/2/index
).
- Convert existing TSM-based shards (or rebuild TSI Preview shards) to support TSI.
- When TSI is enabled, new shards use the TSI index. Existing shards must be converted to support TSI.
- Run the
influx_inspect buildtsi
command to convert existing TSM-based shards (or rebuild TSI Preview shards) to support TSI.
Note: Run the
buildtsi
command using the user account that you are going to run the database as, or ensure that the permissions match afterward.
Step 6: Restart the influxdb
service.
Restart data node
sysvinit systems
service influxdb restart
systemd systems
sudo systemctl restart influxdb
Step 7: Restart traffic to data node
Restart routing read/write requests to the data node server (port 8086) via your load balancer.
If this is the last data node to be upgraded, proceed to step 7. Otherwise, return to Step 1 of Upgrading data nodes and repeat the process for the remaining data nodes.
Step 8: Confirm the upgrade.
Your cluster is now upgraded to InfluxDB Enterprise 1.6.
Check your node version numbers using the influxd-ctl show
command.
The influxd-ctl
utility is available on all meta nodes.
~# influxd-ctl show
Data Nodes
==========
ID TCP Address Version
4 rk-upgrading-01:8088 1.6.6_c1.6.6 # 1.6.6_c1.6.6 = 👍
5 rk-upgrading-02:8088 1.6.6_c1.6.6
6 rk-upgrading-03:8088 1.6.6_c1.6.6
Meta Nodes
==========
TCP Address Version
rk-upgrading-01:8091 1.6.6_c1.6.6
rk-upgrading-02:8091 1.6.6_c1.6.6
rk-upgrading-03:8091 1.6.6_c1.6.6
If you have any issues upgrading your cluster, please do not hesitate to contact InfluxData Support at the email address provided to you when you received your InfluxDB Enterprise license.
Was this page helpful?
Thank you for your feedback!
Support and feedback
Thank you for being part of our community! We welcome and encourage your feedback and bug reports for InfluxDB and this documentation. To find support, use the following resources:
InfluxDB Cloud and InfluxDB Enterprise customers can contact InfluxData Support.