4.85It comes with fixes of 4.84 together with some small enhancements
- VMware: trend graphs for memory and VM/Resource Pool/Cluster total had been added
- VMware: added NFS based datastores
- IBM Power heatmap: unused CPU pools have been in the map, they are excluded now
- IBM i (AS400) beta support
4.84It comes mainly with VMware enhancements.
- Support of cluster stats on VMware vCenter 4.1
- Speed up of back-end processing (parallelization when retrieving data)
- Many other smaller enhancements and fixes.
- OS agent supports Linux generally. It allows to load performance data from any Linux box.
All will be placed under "Linux" item in "SERVER" menu.
- Heatmap: enhanced memory, RED is reported if memory usage is over 90% and paging is over 50kB/sec (in or out).
- NMON data support in historical reports
- New filtering implemented for all tables and all their columns include legends on aggregated graphs
(point mouse on the shadow strip under each table header)
IBM i OS agent is coming soon.
4.81-004Cumulative patch. Install it on the top of version 4.81. Download link.
- OS agent response time data has been added into "Historical reports"
- links from "LPAR search" can be wrong (they might be dirtected to already removed LPARs)
- space characters caused a problem in server, lpar and cluster names
- backlinks from legends & heatmaps fixed
- historical report server - enable tabs for NMON agent data, add resp_time tab
- "load.sh html" can crash with "no space" in very big environments, it is caused by a new added code in 4.80, fixed now
- heatmap crashing when any LPAR has this setting: curr_proc = 0
- CPU pool trend graphs issue: yellow part (borrowed) was wrong (uin fact there was entitled value)
- Refresh icon on LPAR graphs disapeared
- deactivated HMC are excluded from the "HMC total" menu
4.81Bug fixes of 4.80
- OS agent fix for reporting of fcs interfaces dedicated to NPIV and tapes (upgrade the agent to 4.81)
VMware workaround of a VMware bug "BUG PR no:1467231" (affected versions are VCSA 5.5U3 and VCSA 6.0U1 without any patch)
It caused gaps in cluster/resource pool and datastore data
- Historical reports form might fail to load
- Historical reports: aggregated CPU graph did not work
- switch to the HMC view might fail the UI
- SAN graphs: the first interface has not been stacked in aggregated graphs
- CPU workload Estimator did not work in "free" product release
- VMware: VM and datastore names with a space inside caused problems
New major enhancements:
- VMware support
CPU Workload Estimator is enhanced about SMT based rPerf estimations
It gives more accurate estimations based on SMT setting of the target and source systems
Custom Groups are enhanced about "set of custom groups"
It allows grouping of custom groups in sets of them what might be useful especially when you have many of Custom Groups defined
Heatmaps feature has been introduced for CPU and Memory (IBM Power & VMware, LPARs/VMs and servers)
OS agent (upgrade to 4.80 version)
- new metric: response times per FCS and vSCSI adapters
- vSCSI adapter statistics (IO, data throughput, response time) can be found under SAN tabs in the UI
- Alerting thresholds can be specified in "%"
- Dashboard sharing: possibility to export dashboard setting and import it whenever else