Post by hurricanemaxi on Jan 6, 2012 2:49:38 GMT -8
Over a year ago, we presented a list of the top 10 worst practices that you should avoid if you want to maintain the performance and uptime of your Microsoft (Nasdaq: MSFT) Exchange email system.
Since then, some things have changed, while others have not. So here's a revised Top 10 List of "Don'ts" to point out which practices still hold true and which do not.
First, let's look at what has not changed.
1. Deploy JBOD Storage Without RAID: Storage strategies have changed very little. The use of JBOD is not a one-size-fits-all strategy, so when considering this approach, be sure to weigh all of the risks and costs. In most cases, JBOD does not provide a high risk/reward return on investment.
2. Use Third-Party High-Availability Solutions: This has not changed for Exchange or for any other Microsoft products. From a best-practices standpoint, if a product comes with a built-in high-availability feature, IT admins should stick with that feature set. Again, this goes back to risk and reward. Keeping things as simple as possible will present the lowest risk and highest reward.
3. Stretch Your Data Centers for Disaster Recovery: This point has changed very little. Be sure to weigh all of the risks and costs before jumping to a fully redundant disaster-recovery model. As stated in the earlier post, a well-written, tested and executed DR plan will pay off without the complexity of stretched data centers.
4. Build Exchange for Five Nines of Availability: No change here. Exchange is not capable of five nines.
5. Use Fibre Channel: Network convergence is happening, as predicted by many. Running storage Test Drive the Public Cloud for $1. Windows & Linux Cloud Hosting. Click Here. protocols and network protocols over the same network infrastructure is becoming more common. iSCSI will be the primary storage transport of mid-range applications in the future. Exchange has been the leader so far in using iSCSI, but other applications are likely to follow, especially as 10Gbe switching gets cheaper.
6. Deploy Non-Supported Solutions: Vendor support is key with all products, not just Microsoft's offerings. Keeping deployments as simple as possible and deployed in a supported and documented way will again lower risk and increase reward.
Now here's what has changed with respect to the following "Don'ts" and what you should do about each one.
7. Run Without Backups: This is still not a recommended practice in most cases.
What's changed in the last year are backup intervals and their retention periods. A year ago, most backups were performed once or twice a day; more recently customers have chosen to increase the backup interval to four or more times per day (three incremental backups and on full backup per day).
Mercedes Benz HID
rita ranch tucson
Since then, some things have changed, while others have not. So here's a revised Top 10 List of "Don'ts" to point out which practices still hold true and which do not.
First, let's look at what has not changed.
1. Deploy JBOD Storage Without RAID: Storage strategies have changed very little. The use of JBOD is not a one-size-fits-all strategy, so when considering this approach, be sure to weigh all of the risks and costs. In most cases, JBOD does not provide a high risk/reward return on investment.
2. Use Third-Party High-Availability Solutions: This has not changed for Exchange or for any other Microsoft products. From a best-practices standpoint, if a product comes with a built-in high-availability feature, IT admins should stick with that feature set. Again, this goes back to risk and reward. Keeping things as simple as possible will present the lowest risk and highest reward.
3. Stretch Your Data Centers for Disaster Recovery: This point has changed very little. Be sure to weigh all of the risks and costs before jumping to a fully redundant disaster-recovery model. As stated in the earlier post, a well-written, tested and executed DR plan will pay off without the complexity of stretched data centers.
4. Build Exchange for Five Nines of Availability: No change here. Exchange is not capable of five nines.
5. Use Fibre Channel: Network convergence is happening, as predicted by many. Running storage Test Drive the Public Cloud for $1. Windows & Linux Cloud Hosting. Click Here. protocols and network protocols over the same network infrastructure is becoming more common. iSCSI will be the primary storage transport of mid-range applications in the future. Exchange has been the leader so far in using iSCSI, but other applications are likely to follow, especially as 10Gbe switching gets cheaper.
6. Deploy Non-Supported Solutions: Vendor support is key with all products, not just Microsoft's offerings. Keeping deployments as simple as possible and deployed in a supported and documented way will again lower risk and increase reward.
Now here's what has changed with respect to the following "Don'ts" and what you should do about each one.
7. Run Without Backups: This is still not a recommended practice in most cases.
What's changed in the last year are backup intervals and their retention periods. A year ago, most backups were performed once or twice a day; more recently customers have chosen to increase the backup interval to four or more times per day (three incremental backups and on full backup per day).
Mercedes Benz HID
rita ranch tucson