If you click the Sonus category here you’ll quickly notice that I’ve written about the Sonus/Ribbon REST interface on and off for years now, as it’s such a handy way of peeking and poking into the SBC without interacting with the browser.
And so it came to pass last week that when presented with a challenge, I’ve again resorted to REST and PowerShell to deliver the fix.
The challenge – refresh the AD Cache
My customer is planning on migrating a couple of thousand users to Skype for Business quite gradually, and for that we’ve proposed the fairly standard upstream model, with the SBC doing AD lookups to determine if a user is enabled for SfB and thus decide where to send the call. If you’re not familiar with it I drew a picture and added an explanation at the top of this post back in 2014.
The catch is that the SBC’s minimum refresh period is an hour, and the customer doesn’t want to be potentially waiting that long for it to kick in after a user’s been migrated. They’re also big users of automation, and so logging into the SBC to clear the cache by hand isn’t really an option.
Thankfully Sonus added the option to clear the cache to their REST interface, and it didn’t take me a lot of effort to “save-as” a script I’m working on (more on that soon) to come up with “Update-RibbonADCache.ps1”.
If you just run the script on its own, it will prompt you for the FQDN and credentials, then go about its business:
PS C:\> .\Update-RibbonADCache.ps1 About to login SBC FQDN : mysbc.greigin.sydney REST login name : REST REST password : Pa$w0rd Login successful Refresh of the Cache requested DomainController : davros.greigin.sydney BackupStatus : Backup Not Applicable CacheStatus : Cache Not Applicable ID : 1 ADStatus : AD Up DomainController : davros.greigin.sydney BackupStatus : Backup Successful CacheStatus : Cache Active ID : 2 ADStatus : AD Up PS C:\>
In the above results you’ll see Davros (yes, he’s one of our DCs here) listed twice. That’s because he exists in the SBC’s config twice, once for Authentication and again for Call Routing – and that’s why one instance reports a status of “Cache Not Applicable”.
If you give it the lot in the one command it will do the deed, then query the SBC for an update:
PS C:\> .\Update-RibbonADCache.ps1 -SkipUpdateCheck -SbcFQDN mySweLite.greigin.sydney -RestLogin REST -RestPassword Pa$w0rd About to login Login successful Refresh of the Cache requested DomainController : davros.greigin.sydney BackupStatus : Backup Successful CacheStatus : Cache Active ID : 1 ADStatus : AD Up PS C:\>
The script outputs a PowerShell Object, so you can use it in downstream tests, or pipe it to “ft” to customise the output display:
DomainController BackupStatus CacheStatus ID ADStatus ---------------- ------------ ----------- -- -------- davros.greigin.sydney Backup Not Applicable Cache Not Applicable 1 AD Up davros.greigin.sydney Backup Successful Cache Active 2 AD Up
If you add the “-QueryOnly” switch to the above it won’t trash the cache, just query the status to check it’s OK. If you have some kind of automated health checks, this might be a good one to add to your schedule!
It logs too!
With his generous consent I’ve
stolen included Pat Richard’s logging function, so you’ll find relatively detailed logs – sans password though, naturally – in the /logs/ folder it creates where the script lives.
Running it with the extra -verbose or -debug switches I’ve included will spray more info to screen if you’re needing some assistance debugging it, but hopefully you won’t encounter too many problems beyond the obvious issues with typos in the FQDN or bad REST credentials.
Did it Work?
The SBC’s Alarm/Event History is another way you can confirm the script reset the cache, and if you’re automating it this will be an effective human-viewable way of keeping tabs on it.
You can also have the SBC send these events as SNMP traps to your NMS. And no I don’t know why my DC was uncontactable at midnight. Puzzling.
I’ve added update-checking component so it will let you know as updates become available. You can suppress the update check by running it with the “-SkipUpdateCheck” parameter, which you should remember to add if you’re running the script via a scheduled task or some other automated/unattended means.
You’ll find a code-signed version of the script on GitHub. You’re welcome to pinch, adapt or improve upon the code with my blessing. If you encounter any problems with it please create an issue on the repo.
6th June 2020. Updated the Download link to point to GitHub.
1st August 2018. This is the initial release.