hi team, has anyone used local LLMs ollama with ONTAP MCP? Im only getting genetic responses as agent is not giving any spefic information about my cluster. if I used GitHub Copilot, the MCP server works really well.
#Ollama + ONTAP-MCP
plain steeple
drifting ivy
hi @plain steeple I've not tried it, but this issue implied that someone got it working https://github.com/NetApp/ontap-mcp/issues/53 Perhaps adapting these instructions would help https://mcpfind.org/blog/mcp-servers-ollama-local-llm
junior marten
I'm working on some stuff in my homelab, extending the capabilities of the current MCP and working with claude + a few diff local LLMs to see if I can get it all wired up. Waiting on some cabling, so paused at the moment, but so far I've gotten claudecode to load it and see the FAS2750. More soon (tm), as I'll definitely want to PR this work back up to Grinds & team
lone lagoon
Colleague of mine is already using Claude to ask detailed stuff in ONTAP. No MCP server needed he says, Claude is intelligent enough to know about ONTAP APIs. 😮
But Im still an AI-noob can't really wrap my head around when using a MCP server is really necessary.