Shield Drives New Features into HA4i
September 20, 2011 Alex Woodie
The low-cost IBM i high availability solution from Shield Advanced Solutions has gotten more advanced. The latest PTF for HA4i–which became available again last week after an initial delay–introduces new features, such as an enhanced file synchronization process, a new audit process, better handling of object and spool file locks, role swap enhancements, and better management of job logs.
HA4i (formerly named RAP) is a budget-minded high availability solution based on IBM‘s remote journaling technology. The software allows customers to replicate their data and select IBM i objects from a source to a target system, and to conduct role swaps in the event of an emergency. Rounding out the offering are tools for managing aspects of the HA environment, such as file synchronization and automating audits. HA4i may not offer all the bells and whistles of more expensive HA solutions, but it offers enough functionality to get the job done for smaller shops.
Shield has enhanced several components of HA4i with PTF 4, which was first released back in July and was re-released last week after some bugs were addressed.
The Sync Manager saw several enhancements, including smarter handling of object locks. If an object is locked when Sync Manager tries to save it and transport any changes to the target system, Sync Manager will add it back to the queue and try again later, instead of dropping save attempts. Objects that are constantly locked will trigger a delay in the file synchronization job, so it doesn’t consume too much system resources. Also, Sync Manager’s own locks are shorter, Shield says.
Synchronization jobs can take a long time, so Shield now shows the user the progress of the synchronization, how much data is left to send, what objects are currently being replicated, and what its status is (such as locked, or waiting). HA4i also now compresses objects, which can shorten synchronization. Tape saves can also now be used to synchronize objects manually, which will come in handy when customers have very slow network connections.
HA4i now offers new commands that automate the retry processing of object and spool file replication, and better handling of exceptions. The software does its best to fit retries around existing replication jobs, so as not to impact system resources. Also, HA4i offers better filtering of the failure log for objects or spool files that could not be replicated. Now, only one record is added per object, even if the object is constantly changing.
Better handling of job logs is another new feature in this release. To keep job log entries from becoming unmanageable, HA4i will now automatically restart the job log process every 24 hours, which Shield says will keep job logs to a manageable size. Shield has done more weeding with this release, and the product will self-filter unnecessary messages. Running the product in debug mode makes them visible again.
Shield has also upgraded the save and restore process. During a restore, HA4i will now check for the existence of an object, and take appropriate measures to ensure that the latest version is in place. (Shield had run into problems with the IBM i save and restore process ignoring object attributes when the object already exists.) Shield has also added a new process that ensures a physical file can be deleted while the logical files are maintained.
HA4i’s auditing process has also been enhanced to provide a more thorough audit of the replication process. Shield says it’s added a new audit process that automatically applies the attached receiver’s data before the audit is started. Also, source and data files can now be isolated in the audit, which Shield says provides a more granular approach to file auditing.
Shield has also bolstered HA4i’s role swap process. Specifically, users can now initiate the role swap on the source system, which will then execute the role swap command on both the source and the target system. This gives the user more flexibility in restarting the role swap process on either end of the chain, if needed. Previously, the role swap command had to be issued to both the source and target system. Also, Shield added exit points that allow users to call other programs as needed during the role swap.
This release of HA4i also brings a new self-check feature. The software will kick off a job, via the IBM i job scheduler, to make sure that all HA4i processes are running correctly. If something is amiss, it will send a message to the administrator via QSYSOPR and email.
In his Simply “i” blog Shield president Chris Hird says the next release of HA4i will feature major changes to the save and restore process used by HA4i. In short, Shield will write its own save and restore process, although it will still give HA4i customers the option of using IBM’s.
“The APYJRNCHG command does a great job but there are times when some of the features affect the customer experience and IBM is either unwilling or unable to effect a change that would make the process more acceptable,” Hird wrote in his blog in July. “We still have a long way to go with the next release but the early testing of the new apply process is going well.”
For more information on HA4i, see Shield Advanced Solutions’ website at www.shield.ca.