Anyone out there using Adagio Cloud who has had to recover from some kind of failure and wants to share the recovery experience?

I just had an unrecoverable error when I was doing my end of day verification and had to restore from the day before which left me having to re-create an entire day of work.

I'm wondering if the Cloud solution would improve that recovery process or if it would be the same, just with my back up files in a different location...

Thanks in advance for your thoughts!

Hi, Hugh,

Whether you are working on cloud or on premise, if you are using a v9.3A or higher version, your Adagio Backup Options can be set to automatically or to prompt you to backup 'prior' to running a function such as Day End, posting a batch, doing period end, etc..

In the cloud, we also have nightly backup procedures whereby a site's data is backed in case something should go wrong and a restore become necessary. When working on premise, it is also recommended a site set such procedures up. Adagio DataCare may be used to help accomplish this.

If you find you're in a situation where doing a restore is necessary, it's probably a good idea to contact your Adagio consultant for their assistance.

So there's no improvement in the recovery process over the in-house version? I'm still restoring from a backup and redoing the work?

I hear what you're saying regarding backups. We're so gunshy now we don't post during the day. We wait until the end of the day, do a data integrity check/backup and then do the day's postings. However the situation I was just in was that something had gone flaky during the day and we didn't know about it until we ran the end of day integrity check leaving us having to restore from the previous night and re-do the current day's work. So if I understand correctly the Cloud version won't assist me with that situation?

Hugh:

This is what the Support Technicians are telling you.
Select the option in V9.3 to Backup & Post together.
Then, when you do post, you actually have a working backup of the data from that moment.
So, if the data gets damaged or something unexpected happens, then you can just restore the data from the moment just before the post.

You can cycle through many versions of the backups, and you do not have to restore all of the Adagio data if just one module is messed up.
Ask your consultant to give you a heads-up on this feature.

Sam,
Thanks for the input... However the point is that even if you force backups before posting you don't know you need to restore until you've done a verification and seen that it failed. At that point you have lost whatever work was done all the way back to the last backup that didn't contain the error.

So our current protocol is to run verifications before every posting so that we only have restore from the backup before. Otherwise if you did an end of day verification and found that something was damaged during that day, you'd have to incrementally restore each backup and test to see if you'd gone far enough back to be before the error occurred.

The problem with this is that it is time consuming to run a verification every time you post. And furthermore, regardless of how frequently you back up, doing a restore requires you to both identify and recreate your work since the last valid backup. The redoing of the work is just unfortunate, but the bigger issue is identifying what needs to be redone since you could be missing information.

*sigh*

Hugh,

I think the more important question here is why are you getting data integrity errors, that can't be rebuilt, this frequently? What type of errors are they?

Because if the errors you're getting are being caused by hardware or network issues, then Cloud would certainly change that because the data processing is self contained on a server and eliminates most LAN or environment issues.

If we know WHY you're getting the errors, then perhaps we can recommend a different approach.

Do you use the DataCare module? If so, you could schedule it to run several times a day (or even on the hour) with the auto-email function turned on to send you the results for failure or success. Then you would be able to deal with the issue right away. DataCare is included in Cloud Core bundle.

ya, unsure what the issue may be, or if they are related. I've had 2 unrecoverable data errors over the past 3 years and one error that was fixed by installing a 'hot-fix' so it's not an epidemic but....

not sure how DataCare works... if it's operating in the background and doesn't require people to log off or stop work to get an accurate picture that might be of assistance. I'll dig into it. Thanks!

There is an option to run DataCare without requiring exclusive access. However, from the help file:
"If you don't require exclusive access, false errors could be reported if another user has the database open. For that reason, you cannot rebuild unless you have exclusive access."
So you may have to review the errors and run again with exclusive access if it's not obvious from the report.

I'm sure your consultant can demonstrate the many options/benefits of DataCare.