Getting Data In

Identify number of deployment server clients?

Jnewman28
Explorer

When looking to update a deployment enterprise instances, we are exploring using a deployment server.

Splunk documentation says the below:

"When provisioning a deployment server, note the following:

  • If it will be deploying to more than 50 clients, the deployment server must run on a dedicated Splunk Enterprise instance. The instance cannot double as an indexer or search head."

Is this below command the way to view the number of clients?

If so, I have 46, but I am looking to confirm, to see if we need to use a dedicated Splunk enterprise instance for updating using a deployment server.

 

 

 

index=_internal source=*metrics.log group=tcpin_connections splunk_server=* 
| eval host=if(isnull(hostname), sourceHost,hostname) 
| search (host=*) AND (host!="(ALL)")
| eval connectionType=case(fwdType=="uf","univ fwder", fwdType=="lwf", "lightwt fwder",fwdType=="full", "heavy fwder", connectionType=="cooked" or connectionType=="cookedSSL","Splunk fwder", connectionType=="raw" or connectionType=="rawSSL","legacy fwder") 
| eval Ver=if(isnull(version),"pre 4.2",version) 
| fields connectionType sourceIp host kb tcp_eps tcp_KBps splunk_server Ver
| stats min(_time) as first_time, max(_time) as last_time, max(tcp_KBps) as max_tcp_KBps, avg(tcp_KBps) as avg_tcp_KBps, avg(tcp_eps) as avg_tcp_eps, sum(kb) as sum_kb by host sourceIp Ver 
| dedup host
| eval sum_mb = round(sum_kb/1024,0) 
| fields - sum_kb 
| eval avg_tcp_KBps = round(avg_tcp_KBps,0)
| eval max_tcp_KBps = round(max_tcp_KBps,0)
| eval class=case(
avg_tcp_KBps>0 AND avg_tcp_KBps<=128, "Standard Hosts",
avg_tcp_KBps>128 AND avg_tcp_KBps<=256, "Better Hosts",
avg_tcp_KBps>256, "Special Hosts")
| convert ctime(*time) 
| rename first_time as "First seen", last_time as "Last seen", avg_tcp_KBps as "AVG KB/s", avg_tcp_eps as "AVG Events/s per 30 seconds", sum_mb as "Total MB", max_tcp_KBps as "Peak Kbps" connectionType AS "Forwarder Type" sourceIp as "Source IP" host AS "Host" Ver As "Splunk Version"
| sort - "Peak Kbps"

 

 

Labels (1)
0 Karma

Jnewman28
Explorer

Thanks for the insight and confirmation. We currently have a separate IDX Master that houses the License Manager and feeds it to the indexers.

Are you saying it's a good practice to place the deployment server on the (in our case) IDX Master server?

0 Karma

richgalloway
SplunkTrust
SplunkTrust

Co-locating the DS and CM is something you may be able to get away with, but it's not a good practice.  As the Splunk environment grows and becomes busier, the DS and CM will compete for resources and the indexer cluster will suffer for it.  See "Management / Utility Tier Recommendations" in the Splunk Validated Architectures manual (https://www.splunk.com/pdfs/technical-briefs/splunk-validated-architectures.pdf)

---
If this reply helps you, Karma would be appreciated.
0 Karma

richgalloway
SplunkTrust
SplunkTrust

The search looks reasonable and 46 is an OK number for a DS.  The 50 number is not a hard limit, but a guideline.  A lot will depend on how frequently the clients phone home.  However, it's still a Good Practice to separate the DS from the indexers and search heads.  It's common to co-locate it with the License Manager.

---
If this reply helps you, Karma would be appreciated.
Got questions? Get answers!

Join the Splunk Community Slack to learn, troubleshoot, and make connections with fellow Splunk practitioners in real time!

Meet up IRL or virtually!

Join Splunk User Groups to connect and learn in-person by region or remotely by topic or industry.

Get Updates on the Splunk Community!

From Data to Insight: Announcing the Winners of the Splunk Dashboard Contest

Hi Splunkers, First off, thank you to everyone who participated in our very first From Data to Insight: The ...

Splunk Developers: Construct Your Future at the .conf26 Builder Bar

Calling all Splunk architects, platform admins, and app developers: the site is open, and the blueprints are ...

Quick connection discovery mode for forwarders

When a Splunk forwarder loses connectivity to its indexers, it does not always reconnect immediately. In many ...