Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

Register Log in
  • Congratulations Mike Lewis on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

MEANINGFUL Performance/Load Statistics

Status
Not open for further replies.
neuralnode

neuralnode

Technical User
Sep 12, 2007
59
0
0
PL
Hi Everyone,

Here's the tough one for you:

OS: Solaris 10

OBJECTIVE: meaningfully and accurately show processor, memory and disk usage/load per:

1. process
2. zone
3. SunCluster resource group

Seemingly simple, the task may get a bit complex. For example, an Oracle database will usually start many processes that utilize the same shared memory area. However, I'm not really able to say if out of, say, 25 oracle processes, 18 use the same chunk of memory, while 7 use some additional. If this would be the case, I'd have to add the first shared chunk (used by the 18 processes) to the other chunk (used by the remaining 7).
How do I know how memory is being utilized by (i.e. exclusive vs shared) by a bunch of oracle processes?

Somewhat similar question arises in case of SunCluster. Is there a way of telling how different cluster resource groups utilize memory/processor/disk?

Same thing relating to zones: is the command prstat -Z trustworthy?

Any ready tool, set of tools, or even DTrace script that would attain said objective?

Thanx in advance for your input!

--
 
Sort by date Sort by votes
The "[tt]ipcs[/tt]" command can give you useful details about shared memory usage. It can show you the size of the shared memory segment and how many processes are attached to it.
Code: 
# ipcs -micb
IPC status from <running system> as of Thu Mar 24 09:44:56 PDT 2011
T         ID      KEY        MODE        OWNER    GROUP  CREATOR   CGROUP      SEGSZ ISMATTCH
Shared Memory:
m 2113929323   0x100daeb  --rw-rw-r--     root     root     root     root       4192        0
m 2113929322   0x100dae7  --rw-rw-r--     root     root     root     root       4192        0
m 1241513984   0xca002    --rw-rw-rw-     root     root     root     root     464172        0
[b]m   16777253   0xc63cec   --rw-r-----   oracle      dba   oracle      dba  650117120       26[/b]
m         43   0x100d9bc  --rw-rw-r--     root     root     root     root       4192        0
m         42   0x100d97a  --rw-rw-r--     root     root     root     root       4192        0
m         40   0x100d91d  --rw-rw-r--     root     root     root     root       4192        0
m         39   0x100d96e  --rw-rw-r--     root     root     root     root       4192        0
m         34   0x100d8e5  --rw-rw-r--     root     root     root     root       4192        0
m         33   0x100d51d  --rw-rw-r--     root     root     root     root       4192        0
See the man page for options.


 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

neuralnode
  • Locked
  • Question
Monitoring IO performance
Replies
5
Views
87
tarn
tarn
kater96
  • Locked
  • Question
massive bad performance on a sun sparc versus x86 platform
Replies
2
Views
62
SamBones
SamBones
yoojamu
  • Locked
  • Question
How to monitor and compare performance of same command on two servers
Replies
3
Views
91
SamBones
SamBones
uuperl
  • Locked
  • Question
Solaris 10 update 8 upgrade slow mysql performance
Replies
0
Views
53
uuperl
uuperl
kaushikbramesh
  • Locked
  • Question
Appaling performance on Solaris 10 running on M5000 Server
Replies
4
Views
77
blarneyme
blarneyme

Part and Inventory Search

Sponsor

Back
Top