Open Worksheets may be saved for at least these reasons:
To create custom performance views which display statistics of interest.
To investigate performance events for later analysis. A worksheet may be paused on a particular event and then saved, so that others can open the worksheet later and study the event.
The following properties are stored for saved worksheets:
Mouse over saved worksheet entries to expose the following controls:
Single click an entry to open that worksheet. This may take several seconds if the worksheet was paused on a time in the distant past, or if it spanned many days, as the appliance must read the statistic data from disk back into memory.
Worksheet maintenance actions are available under the analytics worksheets context. Use the show to view the current saved worksheets:
walu:> analytics worksheets walu:analytics worksheets> show Worksheets: WORKSHEET OWNER NAME worksheet-000 root 830 MB/s NFSv3 disk worksheet-001 root 8:27 event
Worksheets may be selected so that more details may be viewed. Here one of the statistics is dumped and retrieved in CSV format from the saved worksheet:
walu:analytics worksheets> select worksheet-000 walu:analytics worksheet-000> show Properties: uuid = e268333b-c1f0-401b-97e9-ff7f8ee8dc9b name = 830 MB/s NFSv3 disk owner = root ctime = 2009-9-4 20:04:28 mtime = 2009-9-4 20:07:24 Datasets: DATASET DATE SECONDS NAME dataset-000 2009-9-4 60 nic.kilobytes[device] dataset-001 2009-9-4 60 io.bytes[op] walu:analytics worksheet-000> select dataset-000 csv Time (UTC),KB per second 2009-09-04 20:05:38,840377 2009-09-04 20:05:39,890918 2009-09-04 20:05:40,848037 2009-09-04 20:05:41,851416 2009-09-04 20:05:42,870218 2009-09-04 20:05:43,856288 2009-09-04 20:05:44,872292 2009-09-04 20:05:45,758496 2009-09-04 20:05:46,865732 2009-09-04 20:05:47,881704 [...]
If there was a need to gather Analytics statistics using an automated CLI script over SSH, it would be possible to create a saved worksheet containing the desired statistics which could then be read in this fashion. This is one way to view analytics from the CLI; also see Reading datasets.