Not really a question here, but just wanted to make sure people don't upgrade to the 5.5 RC in their production environments. I was stupid enough to upgrade and thought that after doing a manual backup that it kept the files (I guess it's set to auto-backup nightly and only keeps 1 day). So I lost my backup - total noob mistake - and am now unable to downgrade.

5.5 brings a lot of potential, but we're having some severe issues. I'm working with engineers to iron out the bugs, and I know they rely on us to provide info. I just want to make sure others out there only put this through to their test environments.

Our environment consists of:

  • Windows 7 x86
  • Windows 7 x64
  • Approximately 5,000 total machines
  • 20 locations, each with their own replication server 

Some known issues I've seen in our environment:

  • Agents show AMP connection but aren't checking in unless forced via RUN NOW script.
  • Performance of GUI is horrible.
  • Replication tab takes approximately 5 minutes to load.
  • FILTERS DONT WORK CORRECTLY. This is the biggest issue. If I specify IP ranges, and computer descriptions, it doesn't work. I've already been on with tech support for 3 hours with this issue and it's being escalated. The filters simply don't work as intended.
  • Replication servers aren't replicating.
  • Inventory view by default shows oldest machines first. Every time I go back to my inventory page, it shows machines that are basically MIA instead of the newest check-ins. No way to change or remember this preference (I can hit filter a couple times and make newest first...but if I go back later, it's back to old check ins)
  • If I do a simple inventory search of "software title" contains "CS6", it will take 15-20 minutes to load. 
  • While the inventory searches or replication tabs are loading/timing out, you can't click anywhere else. In order to use the K1000 interface, you need to close the browser entirely and login again. Loading new tabs just times out.
Again, not complaining - I realize I should've read further into backups before proceeding (not realizing they are overwritten each night). I'm sure my info will be valuable to techs/engineers to help get rid of these bugs. I just wanted to give you guys a heads up in case someone else does what I did. It's almost making my environment unusable at this time. I'm looking forward to a stable 5.5 with the new iPhone App so I can deploy things remotely - just need a lot of bugs ironed out. I thought I read something about it coming out any time now...but with these bugs still around (and their updated changelogs don't show fixes for these), I would suggest not being the first to upgrade upon release :) (or just be smart and keep your backups!)
 
That's All! =)
0 Comments   [ + ] Show Comments

Comments

Please log in to comment

Answers

2

Got this from support about the sluggishness reported on the HelpDesk

 

 The analyst took a look a the Kbox via the tether and found that Patch downloads based on patches detected as missing is marked and the download time is every 30 minutes, the new signature downloads is set to every four hours.

The analyst recommends increasing the download of new signatures to once a day in off hours (Lumension is not releasing patches hourly anyway so this is a waste of recourses) and the download of patches detect as missing to once or twice a day depending on your patch schedule. 

While clicking around in helpdesk and watching the logs, the helpdesk became slow as the patch download time came around. 

Not sure why this is an issue now in 5.5 when it clearly wasn't in 5.4.  The UI was snappy and because I have some settings to check for patches in the new release it makes it unresponsive?  Changing to every 12 hours has improved slightly only 8-10 second delay when loading tickets now.

Answered 09/03/2013 by: petelanglois
Fourth Degree Black Belt

Please log in to comment
2

Patching Handshake Issue - For us this was related to the patcher files not being downloaded to the client computer.

Check the HANDSHAKE.LST file and make sure all of the files listed are downloaded to the correct location.

Win7 - C:\ProgramData\Dell\KACE\patches\HANDSHAKE.LST

WinXP - C:\Documents and Settings\All Users\Dell\KACE\patches

The files should show up in the following location:

Win7 - Files -> C:\Program Files (x86)\Dell\KACE

WinXP - Files -> C:\Program Files\Dell\KACE

If the files don't exist you should find a URL inside the HANDSHAKE.LST file pointing to your Kace server. Copy that URL into the address bar of an explorer window and see if the files exist and are reachable.

In our case we had to kick off a patcher re-sync in the Control Panel this process pulled down the correct file that was missing on the server. Now the clients are able to update the patcher correctly and the handshake is working.

Answered 09/05/2013 by: Dak
Senior White Belt

Please log in to comment
1

Well, it is a RC!

I updated my test env two weeks ago to check the RC.

I was not able to replicate the issues you said, but have others.

- CIR don't work (esp. the CommandShellTextReturn) under Windows. Linux works fine

- Patching has a new Step (Handshake) which breaks the patching (simple solution: delete all patches and redownload)

- different issues with K1000 GO (mainly: it accesses K1/adminui, so all users of the APP need to have rights to access this, also 403 if you don't allow all 3 parts of the app (Inv, Deployment and Helpdesk)

Answered 08/27/2013 by: Nico_K
Red Belt

  • Good to hear about some other issues. Yeah, I know it's an RC - I talked with support before installing and they said it's basically final unless any major issues pop up which they hadn't heard of anything. So I expected maybe a couple small bugs, but nothing of this magnitude that I'm seeing on my box.

    Haven't seen the issues you have either. I have patching disabled at this time, but good to hear of a solution, as we were about to start patching again.
    • The full version is supposed to be released sometime this week.
      • It's gone GA so we're just waiting now to get it.
  • This content is currently hidden from public view.
    Reason: Removed by member request
    For more information, visit our FAQ's.
  • Nico_K: How is your test environment set up? Do you have a vm'd kbox?
    • We bought a VM. The price wasn't bad. That's where we test.
    • K1-K3 as VM, some virtual test machines(W7, W8, Linux, Android), a DC, a CA in future (upgraded to a second host, as my main host was only a Dual Core with 8GB) an exchange server, and my two notebooks, my old file server and unregulary some notebooks from different vendors owned by friends of mine.
  • Wow, that really doesn't sell the upgrade very well. I am hoping the next version will come with 1.21 gigawatts for faster installs ;)
    • The biggest hurdle will be getting the kbox to 88 mph...
      • No really, just find a _REALLY_ high building and pray to Newton
  • I accidentally hid my comment above. The release is available for download today and we need a guinea pig to install it...Bueler?
  • How do you get a VM for a test environment?
    • We contacted our rep. The price wasn't bad at all since you are really paying for the licenses.
  • I'm sure that there will be some installs that work fine - but the amount of issues I'm having, I would definitely make sure you're all backed up, and you copy you backup files off the K1000 immediately so that you can restore with tech support's help if needed.
  • I've had LOTS of issues with 5.5RC. Some have been fixed like the incredibly long amount of time to load the replication tab.
    -The GUI is still slower than 5.4 though. It takes roughly 10 seconds after I hit the computer inventory for me to actually be able to type something in the search box.
    -The Scripted Installs I have had issues with. After the upgrade a number of scripts had ;quot& stuck in the batch files at random places making the scripts not run.
    -My SEIM thinks that KACE is running SQL Injection Attacks now and it going bananas
    -over 11k patch definitions have to be sent to each replication share now incuring 2GB+ of WAN use. I thought this release was supposed to reduce the load on the WAN not increase it. I now have patch definitions for OSX in my Microsoft Windows Environment. Whomever thought this one up needs a severe beating.
    -SUPER chatty communications between the clients and server during the transition. We had clients sending hundreds of MB per hour over T1 WAN links back to the KBOX. This seems to have quieted down but I haven't been able to confirm with our network guy who's out of state training.
    -Patches STILL aren't being pushed to every client. I have 500+ PC's and most are not fully patched with critical patches even though the KBOX thinks they are.
    -I had a strange bug where patches were ignoring patch schedules and just randomly patching. I think this may have been due to the changeover from the old agent to the new but I'm not sure.

    Overall I'm very disappointed with the RC and I'm praying the GA fixes all of these issues. I have been a super loyal customer and promoted KACE everywhere I go. This will all change though if these issues are not resolved. I am not sure I can wait until release 6.
    • 5.4 had a service pack. I'm sure whatever issues come up they'll fix them.
      • I hope so cause I don't want to setup LanDesk
    • Good to hear others have issues. I've been really disappointed with the RC. Although it is an RC and bugs are expected, the things I'm experiencing are huge. The GUI is REALLY sluggish. Depending on what I'm searching or clicking on, it can take 3minutes to almost 20 minutes to just do a simple inventory search. I'm also seeing random machines lose an AMP connection even though they're on and I'm already RDP'ed into the machine.
      • That sluggishness went away when they removed the replication share counters as that was holding it up pretty huge. I just have the 10-15 second wait times right now. I only have 500 clients though.
  • What are the replication share counters? I have about 20 replication shares across 20 buildings - all with a 1gbps WAN. A 12MB file took 3 hours to replicate today, and every time I click the replication tab, it sites for 3-5 minutes before it displays the info. If using IE, it just times out to a blank page - Chrome or FireFox must be used.
    • in 5.4 and early 5.5RC there was a column that had the number of objects syncing and the amount of MB/GB left to push. This caused issues as the page couldn't keep up as so many updates were happening. This was removed and hopefully will be back in 6 as it was a useful column for me so I could quickly see what was going on with the replication shares. Now I have to click in each one to see what it's doing which is a MAJOR pain in the a$$.
  • Thanks for the heads up on the RC guys.
    I was planning to install it, but I think I will wait to hear more. I have a large environment (7000 nodes) and 35 replication sites, and full patching. Having those fire off randomly would be disastrous.
    I would be able to install it on a VM in a test environment, but that wouldn't really be an accurate test, as I'm sure a lot of the problems unfold due to the difference with different environments, agents, networks, etc ...
    • Yup. Wow - large environment you have there. I thought mine was big with 5,000 nodes and 20 replication sites. Just weird that I'm having so many issues and others are having different issues...and of course they released 5.5 today.
  • at last: if the GA is avaiable, wait for 4 weeks and then ask support if unshure, as we then know the issues and have patches if nessesary
    • I'll be getting it today need to get other issues resolved and I'm not opening any more cases for 5.5RC.
  • Real question is....if I have the 5.5 RC installed, should I install the GA? Wouldn't get any worse I'd imagine. I'll be sure to pull down the backups before doing so this time :)
    • How much worse could it get? Make sure you take backups and then pull them from the KBOX.
  • ...of course, I try to do a manual backup now, and the site just crashes. I can't load the WebUI now - just goes to Chrome's error page. I'm assuming something with our backup is messed up as well, since the site was unresponsive when I came into the office today as well. Had to do a manual power off and reboot the machine (can still ping and ftp).

    Time to call tech support in the morning. Was hoping the GA may have fixed some of the issues I have, but can't backup my box in order to try this.

    Just one big headache.
    • The link from the support page isn't working to download you have to manually find in by searching the structure of the Client Bundle link.
      • hmmm what do you mean by searching the structure? Can you just post or send me links to the server & client kbin files?
      • Nevermind - found it while searching the archives.
    • I had the same thing happen to me. The backup takes the KBOX offline for upwards of 20 minutes to do the backup. I called support when I couldn't access the KBOX for 10 minutes after clicking backup. Under 5.4 it was 2-3 minutes not 15-20.
      • I came into the office today and it was back up - which was weird because yesterday it was down. Regardless, I pulled the backup files off my kbox and upgraded to the new GA release. It at least fixed my issues with the replication tab taking forever to load (you must have been on a newer one than me - the 5.5 RC I was on showed the total MB/GB of files needing to replicate. It looks like they disabled it on the GA). However, inventory filtering still has the old machines showing up first, and I'm still testing some other errors.
      • I can't reply to you directly for some reason. I guess the reply only goes so deep. They disabled it with a patch so if you bitched you got the patch during RC. I'm sure there were other patches like that as well. I had a tether open so they could patch at will.
  • Hah. Well I was on support for 3 hours and they never even opened a ticket for it, even though they acknowledged the problem. I asked for a ticket to be opened for general performance/sluggish-ness of the WebUI, and they said OK, but nothing was created. Hopefully a bunch of small patches will come out with release notes on what they improve. Still a few bugs I'm seeing here.
    • You need to learn to bitch louder. And if you say ESCALATE this ticket they are supposed to as well to an Analyst or Engineer or something. If I don't get the satisfaction I want I have a slew of e-mail addresses I can send to get things rolling.

      I've done a lot of "stuff" for Dell|Kace and while Rob was at the helm I had his ear. They like to keep us good customers happy as we're the ones promoting their product.

      Bad press in this instance is not good.
  • Wish I'd have seen this before updating late last week. All of my systems within my patch schedules are returning the 'Handshake Failed' error. The UI is notably more sluggish than 5.4 - especially when working with license assets.

    I've got a ticket open with Kace Support on this, but I'm pretty surprised that these issues made it past RC. I saw the comment above about deleting downloaded patches fixing the handshake error, but I'm hesitant to do that without first hearing back from Kace support... I'm glad I took backups right before updating this.
    • Just wanted to follow up - I contacted Dell support and they didn't seem to have much of any idea what was going on. I mentioned to them that I had read of others with similar patching issues that were able to correct it by deleting their patch repository and re-downloading everything, but didn't want to do that without first talking to them.

      Their response was basically, "uhh, yeah.. I was going to suggest doing that next..."

      The only other thing I have that needs to be looked into is the sluggish UI. It has been most noticeable when working with license assets.
      • When I went to 5.5RC it dumped pretty much 2GB of patch signatures to each of my remote replication shares. Most of that was for OSX patches which I don't have a single OSX machine in my environment. I've complained about this to support but I'm not sure if they are forwarding my dismay onward.

        Is this happening on the main KBOX or just replication shares?
      • I would be very interested in hearing your follow-up on this, if you ended up deleting and re-downloading all your patches to fix this or not.

        I too stupidly deleted my backup after the upgrade was successful yesterday, and now am having all kinds of issues:
        -Loading the Replication page is very slow
        -Inventory page defaults to sorting by oldest check-in time, reverts every time the page is reloaded
        -Attempting to deploy patches manually is giving me a "error 'Handshake Failed'" message
        -The agent provisioning progress is stuck at 0% even though I know a number of my machines have checked in and upgraded to the new agent
        -The agent on one of our servers attempted to upgrade automatically, failed, and then caused the server to reboot every hour on the hour, taking down our phone system 3 times yesterday. Wheeee!

        Guess this will be the last time I push through an update the same week it's released.
  • I noticed a large amount of files pushed out to each of my replication shares as well after updating to 5.5, though I didn't pay too close of attention when I saw them.

    Do you have OSX selected in your patch subscriptions? If that isn't selected, I can't think of any reason why KBOX would be downloading them and replicating them among your shares.
    • No we do not have OSX selected. This is a new feature "working as designed" as I was told from support. So much for reducing WAN traffic and communications which was hyped for this release.
  • I was supplied a hotfix for my ticket on Slow Helpdesk. We are down to 5-6 seconds for load on the full ticket queue down from 30+
  • To follow up with my last post - deleting and re-downloading my patch repository did in fact correct the "Handshake error" patching problem. Now that we got that cleared up, 'Detect' schedules seem to run MUCH faster than they previously did in 5.4 SP1.
  • I upgraded to 5.5 (non-RC last Friday).... many cool new toys, but I have a LOT of work to do now. It turns out there is a "known issue" (KA-258) that didn't appear in the known issues list... if you use a batch file as your managed install command line, it cannot be named with the .cmd extension. I now have to download, rename to .bat, upload about 60 of my install packages (~25%) to make them work again....and then they have to replicate to the mirrors...
    • Ugh that sucks. Don't you have Minions to do that?
      • Lol, I wish... I could call in some favors though... good idea
    • For some of the larger ones, I may change from "install.cmd" to "cmd /c start /w install.cmd" ... seems to work
Please log in to comment
Answer this question or Comment on this question for clarity