We haven't made the upgrade to 5.3 server and client yet. I've read the 5.3 Release Notes and it lists the new features and resolved issues but doesn't expand on the changes. We are on server 5.2 with 5.1 clients and are having issues with .NET 1.1 on some Windows 7 computers (about 300 out of 3000) and client AMP connections are dropping so the upgrade is necessary.

Concerns about upgrading:
  • kbscriptrunner.exe is gone which will impact testing scripts, MI's or forcing a checkin while logged in as a limited user in XP (6000 computers). This is also a training issue for all of our technicians that use this feature now.
  • File locations are changed somehow.
  • MIs may be broken.
  • Smart Labels may need to be adjusted?
  • What else needs to be cleaned up after the upgrade? We don't have the resources to run a test environment to see what the changes will be so will make the plunge and then pick up the pieces.
  • Airwolf hasn't upgraded yet either!
  • 5.2 client was skipped. 5.3 client was released a while ago but pulled until AP1 (Agent Patch 1) came out. Is there an AP2 in the works?
0 Comments   [ + ] Show Comments

Comments

Please log in to comment

Rating comments in this legacy AppDeploy message board thread won't reorder them,
so that the conversation will remain readable.

Answers

0
I won't be upgrading until I restore our current DB to one of my test VMs and go over the entire interface extensively.
Answered 09/21/2011 by: airwolf
Tenth Degree Black Belt

Please log in to comment
0
kbscriptrunner.exe is gone which will impact testing scripts, MI's or forcing a checkin while logged in as a limited user in XP (6000 computers). This is also a training issue for all of our technicians that use this feature now.
File locations are changed somehow.
MIs may be broken.
Smart Labels may need to be adjusted?
What else needs to be cleaned up after the upgrade? We don't have the resources to run a test environment to see what the changes will be so will make the plunge and then pick up the pieces.
Airwolf hasn't upgraded yet either!
5.2 client was skipped. 5.3 client was released a while ago but pulled until AP1 (Agent Patch 1) came out. Is there an AP2 in the works?


1. no solution discovered yet other than force checkin from web GUI
2. File location was changed from C:\Program Files\KACE to C:\Program Files\Dell\KACE (and (x86) for the 64-bit clients)
3. MIs are not broken per-se, they just do not fire on the very first checkin (such as a machine you just imaged) until the next days checkin. =(
4. No labels have needed adjusting
5. Scripts are not able to manually run on my box and some clients do not display current logins as before for whatever reason. That and the MIs are my only concern atm.

Hope this helps!
Answered 09/21/2011 by: Swyfter
Yellow Belt

Please log in to comment
0
Thanks for the update Swyfter. Our check-in and scripting interval is every 4 hours rather than every day. Manually forcing an update is necessary to make things happen quicker. It can take three check-ins to get the end result. First check-in the client gets Smart Labels applied for missing software. Second check-in the scripts and or MIs run. Third checkin the inventory is updated to see the result. We use a scripted check-in process for new computer deployments so they are forced to check-in three times with a pause between each. That process was calling kbscriptrunner.exe before so will need to be re-written for the new method of calling runkbot.

Forcing multiple computers to check-in is still not available in 5.3 under the Action menu, right? That means we will still be using a script pushed with Run Now to multiple computers. After upgrading to 5.3 is there a different template offered for Force Check-In? This command [Launch “$(KACE_INSTALL)\runkbot.exe” with params “4 0”.] will only update the Inventory but not run MIs without adding another step [Launch “$(KACE_INSTALL)\runkbot.exe” with params “6 0”.], right? I want a script that will do the same thing as clicking the web interface button "Force Inventory Update" so does the script have to do both steps or only the first inventory 4 0?

We are still planning to do the upgrade to 5.3 soon and are feeling more confident it will be a smooth transition. We have a hardware box and no time for testing in a virtual environment. It makes me nervous when tech support recommends to upgrade the server but test the new client on a limited group first. I was hoping other users who had made the jump to 5.3 AP1 would be the testers for us, as was the case on the first release of the new client! I'm glad we waited before but now we are being motivated to upgrade to solve problems with our current box instead of wanting to upgrade to get new features.
Answered 09/21/2011 by: RichB
Third Degree Green Belt

Please log in to comment
0

Concerns about upgrading:

  • kbscriptrunner.exe is gone which will impact testing scripts, MI's or forcing a checkin while logged in as a limited user in XP (6000 computers). This is also a training issue for all of our technicians that use this feature now.
  • File locations are changed somehow.
  • MIs may be broken.
  • Smart Labels may need to be adjusted?
  • What else needs to be cleaned up after the upgrade? We don't have the resources to run a test environment to see what the changes will be so will make the plunge and then pick up the pieces.
  • Airwolf hasn't upgraded yet either!
  • 5.2 client was skipped. 5.3 client was released a while ago but pulled until AP1 (Agent Patch 1) came out. Is there an AP2 in the works?



  • Force Checkin can be sone with "runkbot 4 0" ; MI's can be ran with "runkbot 6 0"


  • Directory Structure:

Windows Platform:
App: C:\Program Files\Dell\KACE\ (-or- x86 use %ProgramFiles%)
Data: C:\ProgramData\Dell\KACE\ (use %ALLUSERSPROFILE%)
Macintosh Platform:
App: /Library/Application Support/Dell/KACE/bin
Data: /Library/Application Support/Dell/KACE/data/
Unix Platform (RHEL 3-5,):
App: /opt/dell/kace/bin/
Data: /var/dell/kace/

  • MI's are working as well as ever in 5.3.45497 - make sure you use the service pack release. There were some issues during the beta, but they probably shouldn't have been discussed in forums...
  • Labels haven't changed- you may want to expand on this thought to help us understand?
  • Cblake has updated though! ;) Airwolf has a large and complex implementation- I'd be really careful too. That's what backups are for- if you can't live on 5.3 then support can help you move back to 5.2 if you have backups. If you're on a VM life might even be a bit simpler.
  • Bugs & enhancements are always in the works; otherwise engineering doesn't have much to do. I've used every release since 1.2, and they are all good. I agree 5.3 server and agent gave me more pause than most because of the major architectural changes involved. All that aside; engineering spent more time and effort on testing, coding, QA, and beta with this release because it was so major- to that end we likely caught more than expected and that delayed release; to all of our advantage.

Forcing multiple computers to check-in is still not available in 5.3 under the Action menu, right?

Correct, sadly. But I understand why- if it were that easy then K1's would be totally bogged down and networks crashing- if it's somewhat painful then folks stop and think about what they are doing.

I've seen a lot of chatter in the forums about this being a bad release and such- it's not. It was major and engineering spent a lot of time working out kinks. This one made me a bit nervous too guys, but overall the release that's on the site today is stable and good. While we no doubt will find bugs- you guys also know Kace addresses those so make sure you report them to [email=support@kace.com]support@kace.com[/email] if you find any. That's what support is there for; well that and being generally cool people ;)
Answered 09/22/2011 by: cblake
Red Belt

Please log in to comment
0
I've seen a lot of chatter in the forums about this being a bad release and such
I don't remember seeing anyone complaining that it was a bad release. There were a few key issues that caused negative feedback about the release, but these are not all "technical" issues related to the quality of the release. They are more about the way the 5.3 release has been handled.
  1. The release date was over promised - it was delayed WAY beyond the original slated release date. I'd rather hear December and see it in September than hear April and see it in September.
  2. The first "release" was rushed, and so many obvious bugs were found that it was pulled. Kudos for pulling it, but it shouldn't have been released in the first place.
  3. Bug fixes for 5.2 are rolled into 5.3, thus delaying fixes to several issues while we waited for the "official" release of 5.3. Shops running 5.2 are forced to wait or deal with 5.3 bugs. Neither is a good option, and our management teams don't like this.
That having been said, I have been excited for 5.3 for quite some time. I really look forward to upgrading, but I've got to do tons of testing first - because, as you said, my implementation is ridiculously complex. Post-upgrade issues are not an option. We don't want to trade 3 bugs for 5 different ones.
All that aside; engineering spent more time and effort on testing, coding, QA, and beta with this release because it was so major- to that end we likely caught more than expected and that delayed release; to all of our advantage.
I can appreciate this, but my employer does not. They really don't care how much time and effort went into it if the end result is not 100% polished. I don't mean to sound negative, that's just the reality I have to deal with as the administrator of my company's K1000.

Correct, sadly. But I understand why- if it were that easy then K1's would be totally bogged down and networks crashing- if it's somewhat painful then folks stop and think about what they are doing.
I just wanted to comment on this one. If I want the option of forcing a check-in on all 1800 nodes at once, I should have that option. Network lag is my problem at that point.
Answered 09/22/2011 by: airwolf
Tenth Degree Black Belt

Please log in to comment
0
I agree whole heartedly as always, Andy. Keep doing what you do. You guys pushing back and Kace listening to you [us] is what made Kace great in the first place, and I for one want to see it stay that way.
Answered 09/22/2011 by: cblake
Red Belt

Please log in to comment
0

  • Force Checkin can be sone with "runkbot 4 0" ; MI's can be ran with "runkbot 6 0"




sorry can you detail that? I just upgraded and didn't realise I broke my deployment from the k2000 which utilizes a force checkin from a batch file as the last step.

"%ProgramFiles%\kace\kbox\kbscriptrunner.exe"

is what i had... that's not present anymore.
Answered 09/28/2011 by: dogfish182
Orange Belt

Please log in to comment
0
Instead of "kbscriptrunner.exe", you'll need to call "runkbot.exe 4 0" and "runkbot.exe 6 0". The numbers are command line parameters. I believe the path to the agent directory has changed as well (detailed in an earlier post).
Answered 09/28/2011 by: airwolf
Tenth Degree Black Belt

Please log in to comment
Answer this question or Comment on this question for clarity