CX-500 LUNs

Backup and Recovery
A reader recently asked: Am presenlty Testing the I/o performance on the new CX-500 San From Emc And Hba From Emulex,The LUn Size is 1 TB and when doing the Testing with SQlIO Tool with a I/o block size-8 i get 39.3 MB/sec for Random write requestes but when doing the same with Random Read requests i get 6.04 MB/sec as the output,the read/Write cache is enabled with read cache set to 288 and write to 1198,We have sqlserver Oltp Enviroment with mostlty Read Intensive Operations , iam wonering how can i increase the Read i/o ,The Lun design looks bad we should have made smaller Luns and Shared the Data and Log Files among them ,will i get any performance if i set read cache to low value or turn off the read cache or should i recomend to rebuild the san design to smaller chunks , i am not so well versed in san design , what i want is best perfromance for my sql serverr IO ,Pls Suggest We Have Raid 10 On our San with 4D+4D Raid Group sets ,In this Situtation what is the best Setting For the Read Cache for our Hba Emulex(2GB) Thanks&Regards Saradhi

Answer Wiki

Thanks. We'll let you know when a new response is added.

Smaller lun sizes are always better for things like oracle and sql server. There are a couple of things you should consider when setting this up: 1) performance and 2) how you are going to back it up. You should be able to get stats from the CX500 using navisphere analyzer to adjust your cache settings. Hopefully you’re using fiber channel drives. If you’re using ATA, sql server will only perform reasonably well doing sequential I/O. If random, than it will continue to perform slow. To adjust the cache you will have to go into engineering mode and change the parameters. Your raid groups should be fine.

Discuss This Question: 2  Replies

There was an error processing your information. Please try again later.
Thanks. We'll let you know when a new response is added.
Send me notifications when members answer or reply to this question.
  • Epeterson
    I agree with the smaller luns. I would also seperate data and logs on their own physical devices if possible. You may be able to mix RAID-5 for the data and RAID-10 for the logs to get the best price/performance. The best way to determine cache settings is to determine the locality of reference for your reads. I would set it to the recommended default and then monitor it. If the reads are very random, it may not help much. If there is a lot of sequential reading you might be better off trying to tune the storage system using its cache to reduce physical I/O latency.
    0 pointsBadges:
  • Riaan
    For best random read performance ensure your data is striped as widely horizontally - max no of spindles and compressed as narrowly vertically - 'short stroking'. use only as much read cache as other functions require as by definition random read means you would not expect to find any of these blocks in cache. use as much write cache as possible so that writes are delayed and optimised not to interfere with read capability. when you look at the vertical layout place your active data in adjacent bands/stripes and inactive data further away which reduces average stroke length
    0 pointsBadges:

Forgot Password

No problem! Submit your e-mail address below. We'll send you an e-mail containing your password.

Your password has been sent to:

To follow this tag...

There was an error processing your information. Please try again later.

Thanks! We'll email you when relevant content is added and updated.


Share this item with your network: