User Software and Computing
System Status: SSI Metrics
- Not ganglia but useful: Landscape LPC mainly condor - authenticate with grid certificate
- The SSI Metrics pages are meant mainly for sysadmins. Take note of the time in the upper right, and "green box" around node name means online, not necessarily if the node is under high memory/CPU/or network load.
- Requires FNAL SSO to access the pages
- To access the SSI Metrics links from offsite: use the FNAL VPN client - login with FNAL SSO to download the client, and you will need your Services password to operate the VPN
- Interactive nodes:
- CMSLPC Interactive nodes- accessible onsite at FNAL only - note that the green plots just mean the node is online, doesn't measure load
- CMSLPC Interactive nodes: CPU load for each node- accessible onsite at FNAL only
- CMSLPC Interactive nodes: # Processes for each node- accessible onsite at FNAL only
- CMSLPC Interactive nodes: Memory for each node- accessible onsite at FNAL only
- CMSLPC Interactive nodes: Network for each node- accessible onsite at FNAL only
- Condor worker nodes:
- LPC Worker Nodes - green means online
- LPC Worker Nodes load 2 days
- EOS storage nodes:
- EOS on landscape is the quickest way to see if it's up, check for proper space total, MGM errors, change in replica imbalance, or overload on FUSE access
- EOS file storage nodes, EOS storage has a 10GB network, so if many of these are at 1GB apiece, it may slow down EOS (usually overuse of a single file or a few large files but a user)
- CMSLPC Interactive nodes- accessible onsite at FNAL only - note that the green plots just mean the node is online, doesn't measure load
- CMSLPC Interactive nodes: CPU load for each node- accessible onsite at FNAL only
- CMSLPC Interactive nodes: # Processes for each node- accessible onsite at FNAL only
- CMSLPC Interactive nodes: Memory for each node- accessible onsite at FNAL only
- CMSLPC Interactive nodes: Network for each node- accessible onsite at FNAL only
- LPC Worker Nodes - green means online
- LPC Worker Nodes load 2 days
- EOS on landscape is the quickest way to see if it's up, check for proper space total, MGM errors, change in replica imbalance, or overload on FUSE access
- EOS file storage nodes, EOS storage has a 10GB network, so if many of these are at 1GB apiece, it may slow down EOS (usually overuse of a single file or a few large files but a user)