Hello splunkers
I want to splunk enterprise upgrade planing
My splunk system architecture
- 1 Cluster Master Node
- 3 Search Head an SHClustering
- 24 Indexer an Index Cluster Peer Node
All Splunk Version 7.0.1
if i want to Splunk version rolling upgrade by 7.0.x to 8.0.x
Plan A
- master node upgrade 7.0.1 to 8.0.x
- Search Head node upgrade 7.0.1 to 8.0.x
- Index Cluster peer node upgrade 7.0.1 -> 7.1.x -> 7.2.x -> 7.3.x -> 8.0.x
Is It this correct?
OR
Plan B
- master node upgrade 7.0.1 to 8.0.x
- Search Head node upgrade 7.0.1 to 8.0.x
- Index Cluster peer node upgrade 7.0.1 -> 8.0.x
Is It this correct?
If Plan A is correct, let us know why you should upgrade that way.
Please quick answer.
Thank You!
Hi @kimch330,
as you can read at https://docs.splunk.com/Documentation/Splunk/8.0.2/Installation/HowtoupgradeSplunk , you don't need to pass through intermediate versions, you can directly upgrade from 7.0.x to 8.0.x, so the best solution is Plan B.
The problem is that you cannot perform a rolling upgrade because this feature is available from 7.1.x (for more infos see at https://docs.splunk.com/Documentation/Splunk/8.0.2/Indexer/Searchablerollingupgrade ) and it is not convenient to go for this version because the stop would be in any case.
In any case, remember to enable Python 3 on new versions because version 2.7 is maintained by default and above all to check the compatibility of your Apps with version 8 (for more infos see at https://docs.splunk.com/Documentation/Splunk/8.0.2/Python3Migration/AboutMigration ).
To do this you should download and install the App "Splunk Platform Upgrade Readiness App" from Splunkbase ( https://splunkbase.splunk.com/app/4698/ ).
Ciao.
Giuseppe
Hi @kimch330,
as you can read at https://docs.splunk.com/Documentation/Splunk/8.0.2/Installation/HowtoupgradeSplunk , you don't need to pass through intermediate versions, you can directly upgrade from 7.0.x to 8.0.x, so the best solution is Plan B.
The problem is that you cannot perform a rolling upgrade because this feature is available from 7.1.x (for more infos see at https://docs.splunk.com/Documentation/Splunk/8.0.2/Indexer/Searchablerollingupgrade ) and it is not convenient to go for this version because the stop would be in any case.
In any case, remember to enable Python 3 on new versions because version 2.7 is maintained by default and above all to check the compatibility of your Apps with version 8 (for more infos see at https://docs.splunk.com/Documentation/Splunk/8.0.2/Python3Migration/AboutMigration ).
To do this you should download and install the App "Splunk Platform Upgrade Readiness App" from Splunkbase ( https://splunkbase.splunk.com/app/4698/ ).
Ciao.
Giuseppe
Thank you @gcusello !
i have more question..
When upgrading from peer to peer on Single Site Clustering,
Do I need to upgrade the next peer after the bucket is fully validated?
ex) A Peer Upgrade after the bucket is fully validated --> B Peer upgrade .. --> C Peer upgrade ... ?
Hi @kimch330,
Yes, I prefer to do in this way, even if it will take more time.
Ciao.
Giuseppe
Hi @gcusello
why prefer to do in this way?
this way no written on documentation(Single site clustering).
but, Multi site clustering plan At written on it.
(docs : https://docs.splunk.com/Documentation/Splunk/8.0.2/Indexer/Upgradeacluster )
Please tell me your thoughts 🙂
Hi @kimch330,
I was speaking about waiting for byckets fully validation.
It isn't mandatory but I prefer to perform updates in this way, to be more sure that the data are ok before the stop fro upgrade and there is only one node offline and the Splunk HA features give me tranquillity.
Ciao.
Giuseppe