Hello all,
I have a deployment server (license master) and an indexer cluster with 2 search heads. For now I have added deployment server also as a search head in the indexer cluster. I see no issues and the setup is fine.
Now with an additional search head I am planning to do SH clustering . So I think to use the deployment server (which is also a search head in indexer cluster) as SH deployer ?
Anyone has tried his before? Any help appreciated.
Thanks
Veera
Well you can certainly, but you probably don’t want, depends on the size of your infrastructure, you can run anything on one node, but for production environment’s you should not do it for various reasons
- Search head and deplyoer should be separate
- A deployment server which has to server more than 50 clients should act just as deployment server
- The deployer has to be a separate node
We run all this Splunk instances on one server for testing purposes, but in our planned production environment we will have 30 Indexers, 11 Search heads in two clusters, and then separate on virtual machines:
- the two deployer for both clusters
- a License server
- a few deployment servers, synchronized with rsync from a master
- a cluster master, made redundant with a linux cluster
The Splunk docs have good enough source for questions like this. See the SH Cluster deployment overview page to learn about different component and their requirements.
For requirement specifically for Deployer, see this http://docs.splunk.com/Documentation/Splunk/6.2.6/DistSearch/SHCsystemrequirements#Deployer_requirem...
Well you can certainly, but you probably don’t want, depends on the size of your infrastructure, you can run anything on one node, but for production environment’s you should not do it for various reasons
- Search head and deplyoer should be separate
- A deployment server which has to server more than 50 clients should act just as deployment server
- The deployer has to be a separate node
We run all this Splunk instances on one server for testing purposes, but in our planned production environment we will have 30 Indexers, 11 Search heads in two clusters, and then separate on virtual machines:
- the two deployer for both clusters
- a License server
- a few deployment servers, synchronized with rsync from a master
- a cluster master, made redundant with a linux cluster