Go Premium for a chance to win a PS4. Enter to Win

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1815
  • Last Modified:

Red Hat Linux

Hello,

As there is a process in red hat database (oracle )server kblock which is showing high utilization due to which itr shows high load average . . Kindly please tell me how to deal with this situation

Looking forward to your quick reply
0
prashantchauhan
Asked:
prashantchauhan
5 Solutions
 
Michael WorshamInfrastructure / Solutions ArchitectCommented:
We are going to need alot more information than what was stated. We need to know what version of RedHat you are running, version of Oracle, type of hardware you are running it upon, etc. It might even be nice if you provide a screen shot of the problem so we can investigate it a bit further.
0
 
arnoldCommented:
How big is/are the database/s.
Specs on the hardware: CPU, memory, Storage type SAN, local drives in RAID, etc.
Are you using raw, ocfs/2, asm etc for the storage where the oracle databases reside?  is this a RAC setup?

kblockd would tend to suggest that your system is swapping memory.
0
 
Mark GeerlingsDatabase AdministratorCommented:
If this process is busy swapping memory, it could be because the Oracle SGA (System Global Area) is not properly sized for the hardware (physical RAM).  The SGA size is controllable via an init file for Oracle.  But like the others mentioned, we need a lot more information from you before we can make a specific recommendation.  In addition to the types of information the others requested, can you also tell us if this is the only Oracle database instance on this server, or if there are multiple Oracle instances, and tell us if this server is dedicated to running Oracle, or if it is a multi-purpose server that also has other tasks or programs running that may be competing for memory.
0
Independent Software Vendors: We Want Your Opinion

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

 
GnsCommented:
If the process (along with some of the others lgwr etc) are doing a lot of I/O, or attempting a lot of I/O, it may spend a significant time in state "D". This will (per process in this state) add one (unconditional) point to your load average. But it may be quite normal and nothing to worry about at all. Add to that that a multicore machine would likely "survive" a higher LA, for example a LA of 4 would be optimal on a 4-way SMP/core machine (in the sense that all processors has a run queue of 1 at all times), and the picture may not be that dreary at all.

To be able to tell IF you have a problem, I'd recommend using vmstat, iostat and sar from the linux side of things, to get a grip on how "stressed" your system really is.

From the Oracle side of things, assuming you have a supported version (and not legacy 9i or prior), you should be able to get the advisories in OEM to tell you exactly how stressed (and where) oracle is.

In short: Load average is a really bad performance metric. Don't rely on it for real performance tuning.

Cheers
-- Glenn
0
 
GnsCommented:
... And, to add anotgher nice tool from Oracle, I'd suggest you implement statspack... and do some snapshots over a reasonable workload, so that the spreport can show where you spend time (in the DB), and provide some sane tuning recommendations.

Cheers
-- Glenn
0
 
prashantchauhanAuthor Commented:
many Thanks
0

Featured Post

Technology Partners: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

Tackle projects and never again get stuck behind a technical roadblock.
Join Now