Performance statistics for server jobs
Moderators: chulett, rschirm, roy
Performance statistics for server jobs
I am reading through the Ascential DataStage Server Job Developer's Guide to diagnose job limitations. It states, "This is done via the Tracing tab of the Job Run Options dialog box, select the stage you want to monitor and select the Performance statistics check box."
I know how to get the the Job Run Options but there is not a tab for Tracing. Is there something that needs to be "turned on" to allow this feature?
Brad
I know how to get the the Job Run Options but there is not a tab for Tracing. Is there something that needs to be "turned on" to allow this feature?
Brad
Bring up the online help for Director and look for "Tracing Page in Options Dialogue Box", it has all the details. Basically, it should be there for compiled jobs as long as this hasn't been restricted by someone via the Administrator.
-craig
"You can never have too many knives" -- Logan Nine Fingers
"You can never have too many knives" -- Logan Nine Fingers
Log onto the Administrator as dsadm then check the "Properties" for the project you are having issues with.
Check the "Permissions" tab and make sure the appropriate groups are marked as "Datastage Developer" groups (it defaults to all but someone may have changed it). Also see if the "DataStage Operator can see full log" is checked, which it is (again) by default. Not quite sure exactly how this all ties back to your problem, but that's the first place I'd look for something like this.
FYI: Don't be fooled by the "Tracing" tab saying that Tracing is disabled, it is by default - but that doesn't keep the Tracing tab from showing up in the Job Run dialogue box.
HTH,
Check the "Permissions" tab and make sure the appropriate groups are marked as "Datastage Developer" groups (it defaults to all but someone may have changed it). Also see if the "DataStage Operator can see full log" is checked, which it is (again) by default. Not quite sure exactly how this all ties back to your problem, but that's the first place I'd look for something like this.
FYI: Don't be fooled by the "Tracing" tab saying that Tracing is disabled, it is by default - but that doesn't keep the Tracing tab from showing up in the Job Run dialogue box.
HTH,
-craig
"You can never have too many knives" -- Logan Nine Fingers
"You can never have too many knives" -- Logan Nine Fingers
You are barking up the wrong tree with regards to tracing. If you're looking to diagnose bottlenecks in job designs, this ain't it.
Your friends for tuning are:
Your friends for tuning are:
- 1. Small, modular jobs
2. prstat
3. top
4. glance
Kenneth Bland
Rank: Sempai
Belt: First degree black
Fight name: Captain Hook
Signature knockout: right upper cut followed by left hook
Signature submission: Crucifix combined with leg triangle
Rank: Sempai
Belt: First degree black
Fight name: Captain Hook
Signature knockout: right upper cut followed by left hook
Signature submission: Crucifix combined with leg triangle
prstat?kcbland wrote:You are barking up the wrong tree with regards to tracing. If you're looking to diagnose bottlenecks in job designs, this ain't it.
Your friends for tuning are:
- 1. Small, modular jobs
2. prstat
3. top
4. glance
Glance?
-T.J.
Developer of DataStage Parallel Engine (Orchestrate).
Glance is a product that provides a comprehensive system load:
prstat is Sun 2.8 version of top, a commercially available product of similar nature to glance
These are VALUABLE tools to see how much resources are available on a server, and how much resources your processes are using.
Code: Select all
B3694A GlancePlus C.03.45.00 13:41:22 edwhost01 sun4u Current Avg High
--------------------------------------------------------------------------------
CPU Util S SU U | 30% 20% 69%
Disk Util D D |100% 49% 100%
Mem Util S SU U | 34% 39% 42%
Swap Util U UR | 18% 17% 19%
--------------------------------------------------------------------------------
PROCESS LIST Users= 13
User CPU Util Cum Block Block
Process Name PID PPID Pri Name ( 1600% max) CPU IO Rate RSS On
--------------------------------------------------------------------------------
dsapi_slave 27355 27354 30 ardent 97.7/17.2 1913.5 0.3/ 0.0 40.8mb NONE
oracleiwpro 6416 1 51 oracle 34.5/34.3 62.9 0.0/ 0.0 1.09gb SLEEP
phantom 29807 29797 40 ardent 29.7/28.0 113.1 0.3/ 0.0 11.0mb NONE
ora_p016_iw 1199 1 52 oracle 28.4/ 0.2 43.4 0.0/ 0.0 1.09gb SLEEP
ora_p000_iw 18198 1 42 oracle 15.0/ 1.3 638.8 0.0/ 0.0 1.11gb NONE
ora_p010_iw 1159 1 42 oracle 14.3/ 0.1 27.4 0.0/ 0.0 1.09gb NONE
ora_p024_iw 14788 1 53 oracle 14.1/ 0.3 56.5 0.0/ 0.0 1.09gb SLEEP
ora_p001_iw 18200 1 57 oracle 11.6/ 1.3 645.6 0.0/ 0.0 1.11gb SLEEP
ora_p005_iw 18208 1 55 oracle 9.3/ 1.4 721.5 0.0/ 0.0 1.11gb NONE
ora_p026_iw 14792 1 56 oracle 9.0/ 0.2 37.9 0.0/ 0.0 1.09gb SLEEP
ora_p007_iw 18212 1 53 oracle 8.6/ 1.4 710.4 0.0/ 0.0 1.11gb SLEEP
ora_p018_iw 1218 1 53 oracle 8.4/ 0.2 34.6 0.0/ 0.0 1.09gb SLEEP
Code: Select all
27355 ardent 50M 41M cpu1 0 0 0:32.36 6.0% dsapi_slave/4
12017 oracle 1162M 1142M cpu24 21 0 0:00.10 2.1% oracle/2
6416 oracle 1137M 1117M sleep 32 0 0:01.16 2.0% oracle/1
29807 ardent 20M 11M sleep 0 0 0:01.56 1.4% uvsh/4
18214 oracle 1154M 1134M sleep 32 0 0:05.34 1.0% oracle/11
1199 oracle 1138M 1116M sleep 23 0 0:00.50 0.9% oracle/11
1159 oracle 1138M 1115M sleep 59 0 0:00.34 0.9% oracle/11
18204 oracle 1161M 1141M sleep 58 0 0:35.12 0.9% oracle/11
1211 oracle 1138M 1115M sleep 22 0 0:00.35 0.9% oracle/11
18198 oracle 1157M 1137M sleep 22 0 0:10.45 0.8% oracle/11
14788 oracle 1138M 1117M sleep 58 0 0:01.02 0.7% oracle/11
18216 oracle 1154M 1133M sleep 48 0 0:05.30 0.6% oracle/11
14778 oracle 1138M 1116M sleep 34 0 0:00.52 0.6% oracle/11
1218 oracle 1138M 1117M sleep 48 0 0:00.40 0.6% oracle/11
14792 oracle 1138M 1115M sleep 58 0 0:00.43 0.5% oracle/11
NPROC USERNAME SIZE RSS MEMORY TIME CPU
134 oracle 120G 117G 99% 7:28.23 21%
34 ardent 370M 201M 0.2% 1:41.33 7.5%
3 patrol 18M 16M 0.0% 23:26.09 0.1%
55 root 341M 238M 0.2% 2:43.22 0.0%
12 billingj 55M 40M 0.0% 0:00.04 0.0%
1 daemon 2544K 1200K 0.0% 0:00.00 0.0%
Total: 287 processes, 1335 lwps, load averages: 5.75, 5.04, 3.46
Kenneth Bland
Rank: Sempai
Belt: First degree black
Fight name: Captain Hook
Signature knockout: right upper cut followed by left hook
Signature submission: Crucifix combined with leg triangle
Rank: Sempai
Belt: First degree black
Fight name: Captain Hook
Signature knockout: right upper cut followed by left hook
Signature submission: Crucifix combined with leg triangle
Craig,
I have checked what you suggested and still am left in the same situation. Unfortunately, I have encountered other situations where the DS documentation shows one thing and our DS environment does not reflect what is documented. It is frustrating to say the least. Being a DS novice makes it more so.
Brad
I have checked what you suggested and still am left in the same situation. Unfortunately, I have encountered other situations where the DS documentation shows one thing and our DS environment does not reflect what is documented. It is frustrating to say the least. Being a DS novice makes it more so.
Brad
chulett wrote:Log onto the Administrator as dsadm then check the "Properties" for the project you are having issues with.
Check the "Permissions" tab and make sure the appropriate groups are marked as "Datastage Developer" groups (it defaults to all but someone may have changed it). Also see if the "DataStage Operator can see full log" is checked, which it is (again) by default. Not quite sure exactly how this all ties back to your problem, but that's the first place I'd look for something like this.
FYI: Don't be fooled by the "Tracing" tab saying that Tracing is disabled, it is by default - but that doesn't keep the Tracing tab from showing up in the Job Run dialogue box.
HTH,