I want to configure dedicated deployment server for 50 clients, my deployment server specification is Oracle Linux, 12GB RAM, 8 CPU Cores.
But after all clients are registered to deployment server, only 25 clients detected in Forwarder Management, I try to reload using Splunk reload deploy-server command, for a while 50 clients detected and after that it's back to 25 clients.
The forwarder management dashboard does not look very far back in time. So if your deployment clients are phoning home infrequently, they may not show up in the dashboard. I don't think that you can adjust that dashboard, but you could also look at the Distributed Management Console to see if the clients are sending data, etc.
And here are some searches that you can customize for yourself. I would run the searches for at least the last hour, perhaps longer.
Are apps being downloaded? Also shows when Splunk was restarted:
index=_internal sourcetype=splunkd (component=DeployedApplication OR component=PackageDownloadRestHandler OR
(component=loader start OR restart)) | table _time log_level host app message component | reverse
Is the deployment client phoning home?
index=_internal (*phonehome* component=DC*) OR component=DC:HandshakeReplyHandler OR component=ClientSessionsManager
| reverse | table _time host log_level message component
A couple of more general searches that you can use a starting point:
index=_internal component=metrics group=deploy* sourcetype=splunkd
index=_internal sourcetype=splunkd component=ClientSessionsManager OR component=DC:DeploymentClient OR
component=DSManager OR component=DS_DC_Common OR component=DeploymentServer
Now that you have these searches, you could customize them to help you answer specific diagnostic questions like "When was the last time that a particular client downloaded anything?" or "When was the last time that this app was downloaded by any client?"
Also remember that the log_level field will distinguish between errors, warnings and other message types.