Add the customer’s answers under the questions, then cut/paste everything into an SR note.

This list is available as a Word document that you can download.

This page as well as the attached document are updated by Vinzenz Esser.

• What is your performance expectation?
• What was the original performance and what is the performance now?
• When was degradation detected?
• Was the degradation gradual?
• Is the degradation seen on both reads and writes?
• Does the problem exist on only one client machine?
• Does the problem exist when only one client is accessing the file system or only when it is under full load? 
• Does the problem disappear when other load is removed from the system?
• Does the problem exist on all file systems or just one?
• Do all clients exhibit the same behavior?
• Which application(s) exhibit the performance degradation?
• Do all applications exhibit the performance degradation?
• Do standard tools exhibit the issue? (I.e. dd, lmdd, cvcp)
• Has the application performed satisfactorily in the past?
• What is the IO profile of the application? (Application docs or cvlog on the client may provide help here)

• Were any firmware changes made to SAN/LAN components? (FC or network switches, disk arrays, HBA’s, etc)
• Were any OS changes implemented? (New service packs, major OS upgrades, etc)
• Were any new clients added to the SN environment?

• Were there any changes made to workflows?
• Were any new loads added to the system?
• Has there been any new project data added to the system?

• Are client to/from MDC communications slow? (Latency test from cvadmin)
• Are there latencies reading/writing from/to meta data and/or journal disk? (cvlog)
• Are there latencies reading/writing from/to data disks? (cvlog)
• Is there a high percentage of file fragmentation?
• Is there a high percentage of free space fragmentation?
• Does data go buffer cache or DMA? (application docs and/or cvlog in conjunction with mount options to determine buffer cache or DMA.)
• How full is the file system?