On page 5, the NetSUS 3.0 User Guide provides [https://myhostname.local/webadmin] as the location to access WebAdmin from.

If I attempt to access the equivalent of this path on a server running under CentOS 6.5, the server returns an HTTP 404.

If I attempt to access the server at [https://myhostname.local/] instead, I'm able to access WebAdmin successfully.

Steps to Reproduce

1. On the desired server, run NetSUS under a clean installation of CentOS 6.5. Make sure all available updates have been applied to CentOS with yum.
2. In a Web browser, access the equivalent of [https://myhostname.local/webadmin] on the server used in step 1.
3. If the Web browser returns any sort of warning that the site's security certificate seems suspicious, accept ("allow" or "trust") the security certificate to continue. Once you have done this, note the content of the page that displays in the Web Browser.
4. In the same Web browser, visit the equivalent of [https://myhostname.local/], and note the content of the page that displays in the Web browser.

After completing step 3, I'm observing that the Web browser returns an HTTP 404 ("not found"), where instead I would expect to see the landing page for NetSUS WebAdmin.

After completing step 4, I'm observing that the landing page for the NetSUS WebAdmin displays, where instead I would expect it not to display.

Notes

When I access NetSUS WebAdmin, I enter the server name as an IP address. The IP address is a location on my LAN. So if I had a server on 192.168.0.5, I would use [https://192.168.0.5/webadmin] in step 2 of the STR I provide above, and I would use [https://192.168.0.5] in step 4.

Categorically, it seems like this could be a potential issue for users who plan on using their web root for something other than NetSUS WebAdmin.

I'm only reproducing this under CentOS 6.5. I'm not reproducing this under Ubuntu LTS Server 10.04. I'm not able to test with Ubuntu LTS Server 12.04, due to #24. I do not have a copy of RHEL 6.4 (or later) to test this with.

I reproduced this in Chrome for Mac (version 34.0.1847.131), under OS X 10.9.1. I have not attempted to reproduce this yet with any other Web browsers.

It might also be worth noting that when I reproduced this, CentOS was running in a VirtualBox (v 4.3.10 r93012) virtual machine, with a host machine running under OS X 10.9.1.

Cant install NetSUS on RHEL 6.4. The OS set is setup a bit different than normal.
No Sudoers file�. Root access is granted another way.
I cannot acquire compat-db47 through my current repo. This appears to be where the installer script bombs out.

Here is the installation procedure:

[root@ tmp]# ./NetSUS_3.0.2.run
Verifying archive integrity... All good.
Uncompressing NetSUS Installer............
Starting the NetSUS Installation
Checking installation requirements...
Checking for a supported OS...This system is not registered to Red Hat Subscription Management. You can use subscription-manager to register.
RedHat
Checking for a 64-bit OS...OK
Passed all requirements checking!

Is this a standalone installation?
Answer yes unless you are creating an image of the appliance to deploy in multiple locations

Standalone? (y/n): n

The following will be installed

• Appliance Web Interface
• NetBoot Server
• Software Updates Server

Proceed? (y/n): t
Proceed? (y/n): y
Installing...
Verifying archive integrity... All good.
Uncompressing WebAdmin Installer.......................................................................................................................................................................................................
Starting Web Application Installation
iptables: Saving firewall rules to /etc/sysconfig/iptables:[ OK ]
sed: can't read /etc/sudoers: No such file or directory
webadminInstall.sh: line 145: /etc/sudoers.d/webadmin: No such file or directory
chmod: cannot access `/etc/sudoers.d/webadmin': No such file or directory
Restarting apache...
OK
Finished deploying the appliance web application
Verifying archive integrity... All good.
Uncompressing NetBoot Installer..............................
Starting NetBoot Installation
This system is not registered to Red Hat Subscription Management. You can use subscription-manager to register.
Error: Nothing to do

On the Software Update Server page (in NetSUS WebAdmin), the value Mar 17 displays consistently Last Sync:, regardless of which day today's date is.

Steps to Reproduce

1. Run NetSUS (ideally, on a server that is fully updated).
2. From a Web browser, access NetSUS WebAdmin on the server.
3. From the WebAdmin sidebar, select Software Update Server.
4. On the Software Update Server page, click Sync Manually.
5. If no value appears at first under Last Sync:, refresh the page in the Web browser.

Once I follow these steps, I'm observing that Mar 17 displays consistently as the value for Last Sync:.

Notes

I reproduce this behavior with servers running under CentOS 6.5 and under Ubuntu LTS Server 10.04. I was not able to test with a server running Ubuntu LTS Server 12.04 due to #24.

Every time I reproduced this behavior, I also navigated to Settings > Date & Time and confirmed that the timestamp returned by Date & Time matched the current local system time.

I've tested this each of the last three calendar days. Each day the value that displays is March 17th.

I've reproduced this in Chrome for Mac (version 34.0.1847.131) and Safari 7.0.3 (9537.75.14), under OS X 10.9.1. I have not attempted to reproduce this yet with any other Web browsers. To me it seems like this could be a limited sample group; by testing with more browsers (and on additional platforms), it seems like it should be possible to increase the sample group (and determine whether it might appear that the reproducibility of this behavior is limited by either browser or platform).

It might also be worth noting that when I reproduced this, CentOS and Ubuntu were (each) running in a (separate) VirtualBox (v 4.3.10 r93012) virtual machine, with a host machine running under OS X 10.9.1.

Running NetSUS 3.0.2 with Ubuntu 14.04.1 on NetSUS appliance.

I can checkmark an update in our production catalog and when I click apply the checkmark is gone.

Hope I am not the only one experiencing troubles with 10.9.2.
Netbooting a 10.9.1 was working fine. Now with 10.9.2 I cannot boot "old" machines - just Retina MacBooks. Thank you for your help!

Steps to Reproduce:

1. Configure a SUS branch in the NetSUS v. 3.0
2. Enable one or more of the latest updates from Apple. (Safari 6.1.2, iTunes 11.1.5, and Security Update 2014-001 are all viable candidates.)
   2a (optional). Enable one or more older updates (I used Airport Utility 5.6.1.)
3. In Terminal, curl the branch URL for the branch with the latest updates selected.

curl http://NETSUSIP/content/catalogs/index_BRANCH.sucatalog

Expected Results:

A full list of all enabled software updates is included in the curl results, including both the previous update versions and the latest update versions. All enabled updates are available for eligible clients pointed to the SUS.

Actual Results:

None of the latest software updates are included in the curl output nor are they available to eligible clients pointed to the SUS, only selected earlier update versions are returned.

Edit: my angled brackets in the URL were incompatible with the formatting of this page.

If NetSUS is installed from the current trunk/ main branch (last commit 152370d), the page on WebAdmin for Software Update Server (webadmin/SUS.php) displays several unexpected strings (under the branch list and as the last sync date for Last Sync:). The strings appear to relate to an error messages returned when attempting to execute adminHelper.sh.

Specifically, here is what I'm seeing:

• In place of a date string, the page displays the following message: scripts/adminHelper.sh: 432: Syntax error: "(" unexpected (expecting ";;")
• The branch list displays several unexpected branches, each named with a segment of he error message that displays in place of a date string for Last Sync:.

Steps to reproduce:

1. Install NetSUS under a clean, updated installation of Ubuntu LTS 10.04 Server.
2. Access the WebAdmin page for Software Update Server: https://myhostname.local/webadmin/SUS.php
3. On the WebAdmin Page for Software Update Server, review the branch list and the Last Sync date.

Notes

I'm reproducing this on a VirtualBox VM on host that is running under OS X 10.9.1.

I do not reproduce this behavior under CentOS 6.5. I am not able to test this under RHEL 6.4 (or later).

I am not currently able to test this under Ubuntu LTS Server 12.04 due to #28.

All Yosemite clients fail to install "Security Update 2016-003" (031-51438) with the following error...

Installing Security Update 2016-003
Error downloading Security Update 2016-003: “Security Update 2016-003” could not be verified. It may have been corrupted or tampered with during downloading. Click Update to download and check the update again.
Done.

Error downloading updates.

Considerations:

• Yosemite clients get same error with NetSUS v3.0, v4.0 and 4.1.0
• All NetSUS running RHEL 7.2 (clean installs via .run; no upgrades)
• Manually remove and purge product_id (per reposado reference.md) did not resolve issue.
• Security Update 2016-003 for Mavericks installs successfully (single payload).
• Security Update 2016-003 for Yosemite includes 2 payloads (Firmware.pkg and SecUpd2016-003Yosemite.pkg).

Snippet from /var/log/install.log (showing 2 payloads and fatal error message)...
Jul 11 15:13:36 YosemiteMac.mydomain.com softwareupdated[29587]: Ramped updates marked Jul 11 15:13:36 YosemiteMac.mydomain.com installd[2606]: PackageKit: Adding client PKInstallDaemonClient pid=29587, uid=200 (/System/Library/CoreServices/Software Update.app/Contents/Resources/softwareupdated) Jul 11 15:13:36 YosemiteMac.mydomain.com installd[2606]: PackageKit: Removing client PKInstallDaemonClient pid=29587, uid=200 (/System/Library/CoreServices/Software Update.app/Contents/Resources/softwareupdated) Jul 11 15:13:36 YosemiteMac.mydomain.com softwareupdated[29587]: SoftwareUpdate: Added foreground transaction [0x9] for Security Update 2016-003-10.10.5 Jul 11 15:13:36 YosemiteMac.mydomain.com softwareupdated[29587]: SoftwareUpdate: starting download of 031-51438 (Security Update 2016-003-10.10.5) Jul 11 15:13:36 YosemiteMac.mydomain.com softwareupdated[29587]: ContentLocator: Modified URL is: http://exampleip:55936/content/downloads/15/58/031-51438/qjm273o5c0xevcf1d1t1i50xpdmg3mvofs/FirmwareUpdate.pkg?source=swcdn.apple.com Jul 11 15:13:36 YosemiteMac.mydomain.com softwareupdated[29587]: ContentLocator: Modified URL is: http://exampleip:55936/content/downloads/15/58/031-51438/qjm273o5c0xevcf1d1t1i50xpdmg3mvofs/SecUpd2016-003Yosemite.pkg?source=swcdn.apple.com Jul 11 15:13:49 YosemiteMac.mydomain.com softwareupdated[29587]: Finished downloading package FirmwareUpdate.pkg to file:///var/folders/zz/zyxvpxvq6csfxvn_n00000s0000068/C/com.apple.SoftwareUpdate/CFNetworkDownload_QTx7Sz.tmp (error (null)) Jul 11 15:13:50 YosemiteMac.mydomain.com suhelperd[29588]: Verifying package at path: /Library/Updates/031-51438/FirmwareUpdate.pkg Jul 11 15:14:15 YosemiteMac.mydomain.com softwareupdated[29587]: Finished downloading package SecUpd2016-003Yosemite.pkg to file:///var/folders/zz/zyxvpxvq6csfxvn_n00000s0000068/C/com.apple.SoftwareUpdate/CFNetworkDownload_UEgISR.tmp (error (null)) Jul 11 15:14:17 YosemiteMac.mydomain.com suhelperd[29588]: Verifying package at path: /Library/Updates/031-51438/SecUpd2016-003Yosemite.pkg Jul 11 15:14:17 YosemiteMac.mydomain.com softwareupdated[29587]: No more tasks - invalidating session now Jul 11 15:14:17 YosemiteMac.mydomain.com softwareupdated[29587]: SoftwareUpdate: finished download of 031-51438 Jul 11 15:14:17 YosemiteMac.mydomain.com installd[2606]: PackageKit: Adding client PKInstallDaemonClient pid=29587, uid=200 (/System/Library/CoreServices/Software Update.app/Contents/Resources/softwareupdated) Jul 11 15:14:17 YosemiteMac.mydomain.com installd[2606]: PackageKit: Removing client PKInstallDaemonClient pid=29587, uid=200 (/System/Library/CoreServices/Software Update.app/Contents/Resources/softwareupdated) Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: PackageKit: Adding client PKInstallDaemonClient pid=29587, uid=200 (/System/Library/CoreServices/Software Update.app/Contents/Resources/softwareupdated) Jul 11 15:14:18 YosemiteMac.mydomain.com softwareupdated[29587]: PackageKit: Enqueuing install with framework-specified quality of service (utility) Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: PackageKit: ----- Begin install ----- Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: PackageKit: request=PKInstallRequest <2 packages, destination=/> Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: PackageKit: packages=( "PKLeopardPackage <file:///Library/Updates/031-51438/SecUpd2016-003Yosemite.pkg>", "PKLeopardPackage <file:///Library/Updates/031-51438/FirmwareUpdate.pkg>" ) Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: PackageKit: Skipping component "com.apple.security.gkopaque" (48.0.0-1.0.0-48000000000000-*) because the version 86.0.0-1.0.0-86000000000000-1 is already installed at /private/var/db/gkopaque.bundle. Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: Package Authoring Warning: The deleteObsoleteLanguages flag is ignored. No languages will be removed. Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: PackageKit: Will do receipt-based obsoleting for package identifier com.apple.update.firmwareupdate (prefix path=/) Jul 11 15:14:18 YosemiteMac.mydomain.com installd[2606]: Package Authoring Warning: The deleteObsoleteLanguages flag is ignored. No languages will be removed. Jul 11 15:14:19 YosemiteMac.mydomain.com installd[2606]: PackageKit: Extracting file:///Library/Updates/031-51438/SecUpd2016-003Yosemite.pkg (destination=/var/folders/zz/zyxvpxvq6csfxvn_n0000000000000/C/PKInstallSandboxManager/DA7F3870-415C-4C38-8FE0-875A359E7903.activeSandbox/Root, uid=0) Jul 11 15:15:01 YosemiteMac.mydomain.com installd[2606]: PackageKit: Install Failed: Error Domain=PKInstallErrorDomain Code=110 "An error occurred while extracting files from the package “FirmwareUpdate.pkg”." UserInfo=0x7fdf71a98f80 {NSUnderlyingError=0x7fdf6df92b00 "archive verify failed", NSURL=file:///Library/Updates/031-51438/FirmwareUpdate.pkg, PKInstallPackageIdentifier=com.apple.update.firmwareupdate, NSLocalizedDescription=An error occurred while extracting files from the package “FirmwareUpdate.pkg”.} { NSLocalizedDescription = "An error occurred while extracting files from the package \U201cFirmwareUpdate.pkg\U201d."; NSURL = "file:///Library/Updates/031-51438/FirmwareUpdate.pkg"; NSUnderlyingError = "Error Domain=PKXARArchiveErrorDomain Code=101 \"archive verify failed\" UserInfo=0x7fdf6df6d620 {NSURL=file:///Library/Updates/031-51438/FirmwareUpdate.pkg#Scripts, NSFileOwnerAccountID=0, NSFileHFSTypeCode=0, NSFileSystemFileNumber=19534741, NSFileExtensionHidden=false, NSFileSystemNumber=16777224, NSFileSize=115531136, NSFileGroupOwnerAccountID=0, NSFileOwnerAccountName=root, NSFilePosixPermissions=420, NSFileHFSCreatorCode=0, NSFileCreationDate=2016-07-11 22:13:36 +0000, NSFileType=NSFileTypeRegular, NSFileGroupOwnerAccountName=wheel, NSFileReferenceCount=1, NSFileModificationDate=2016-07-11 22:13:49 +0000, NSLocalizedDescription=archive verify failed}"; PKInstallPackageIdentifier = "com.apple.update.firmwareupdate"; }

The user guide outlines that it should be possible for users to access an AFP share on the NetSUS server.

Under Linux, AFP service requires netatalk. In Ubuntu, this requires that the package netatalk be installed.

When the current 3.0 standalone installer (the coppy of the installer linked to from the front page) is run under Ubuntu LTS 12.04 server, however, netatalk does not appear to be installed as expected.

Steps to Reproduce

1. Start with a machine running a clean installation of Ubuntu LTS 12.04 Server. Make sure all of the packages are updated with apt-get.
2. On the clean Ubuntu machine from step 1, run the NetSUS 3.0 standalone installer linked to from the front page of this project.
3. Attempt to locate netatalk. I did this by running $ dpkg-query -l netatalk. If netatalk is not installed, this command returns an error message that no packages matching netatalk were found; otherwise, it returns name, version number, description, and other information for the currently-installed version of the netatalk package. Another easy way to do this is to look at /etc/netatalk: on a clean system this folder should exist only if netatalk is installed.

Notes

My (oblique) understanding is that it appears NetBoot also requires netatalk (somehow).

I'm not reproducing this in Ubuntu LTS Server 10.04. I'm only reproducing this in 12.04.

I cannot test this in an installer from a copy of NetSUS cloned from this repository (last checkin 152370d) due to #28.

My test machines are VirtualBox VMs, on a host running under OS X 10.9.2.

Shrinking disks is a lot harder than growing - or adding further storage

We are deploying multiple nodes to just use NetBoot (not SUS) and actual disk usage is much much lower than the supposed minimum requirements. Would be good to have some flexibility for admins to be able to remove the huge default and replace with another disk of appropriate size for their installation

Something along the lines of for the default:
/ = 15 GB
/srv/NetBoot = 200 GB

I could easily replace the NetBoot partition with one more appropriate in this case - however I am having a hard time justifying 6 appliances with 1200 GB that will not be used

It would be beneficial to have the version number displayed via web-console so that you know it was updated for security audits.

How do I change how the boot image is sent over to the client from http to NFS? I am trying to use the bless command to remote reboot the client machine into netboot mode but it only works so far with NFS and NetSUS is sending the netboot image using http

After upgrading to NetSUS_3.0.2 i get error awk: line 2: function gensub never defined in the IP and Gateway addresses and netboot will not start

Tested on a fully up to date Centos 7 install. The various catalogs (index-10.9-mountainlion-lion-snowleopard-leopard.merged-1.sucatalog and so on) are not generated by default when the NetSUS is installed via the installer package when choosing to not cache updates.

This issue is not present in the 3.02 OVA,4.0 OVA, or the 3.02 installer, but I am able to duplicate it on Centos 7 using the installer. Manually adding the apple catalogs to /var/lib/reposado/preferences.plist resolves the issue but this probably should still be addressed somehow.

It seems the NetBoot does not work properly on CentOS or RedHat. When trying to boot to the advertised NetBoot image a flashing globe appears on the client screen, after which it just boots to the hard disk. I am able to get it working on Ubuntu with the same image, the only difference being a spinning globe instead of a flashing one and it boots to the NetBoot image successfully. The dhcpd.conf file is exactly the same across OS's and all necessary files have the right privileges.

A unique log message I saw on the CentOS 7 server:
DHCPACK TO

The OS detection for this script is flawed:

if [ -f "/usr/bin/lsb_release" ]; then

ubuntuVersion=lsb_release -s -d

case $ubuntuVersion in
Ubuntu\ 12.04)
detectedOS="Ubuntu"
;;
Ubuntu\ 10.04)
detectedOS="Ubuntu"
;;
esac

elif [ -f "/etc/system-release" ]; then

I have a RHEL 6.5 box which also has /usr/bin/lsb_release as part of the redhat-lsb-core package, so I never get to the elif in this code block and detectedOS is left blank.

There is a fix already for this in base/testOSRequirements.sh but it hasn't been applied here. I'm not sure how people are even using NetSUS 3.0 on a RHEL box with this bug?

I worked around this by manually setting detectedOS in this adminHelper.sh but I'm wondering if there are other places where this OS detection isn't working?

Josh

The case statement in /var/www/html/webadmin/scripts/adminHelper.sh needs
to be updated to get 10.9 updates to propagate to the root branch when
³applying². Add in the second to last entry.

rootBranch)
cp
"/srv/SUS/html/content/catalogs/others/index-leopard.merged-1_$2.sucatalog"
"/srv/SUS/html/index-leopard.merged-1.sucatalog"
cp
"/srv/SUS/html/content/catalogs/others/index-lion-snowleopard-leopard.merge
d-1_$2.sucatalog"
"/srv/SUS/html/index-lion-snowleopard-leopard.merged-1.sucatalog"
cp
"/srv/SUS/html/content/catalogs/others/index-leopard-snowleopard.merged-1_$
2.sucatalog" "/srv/SUS/html/index-leopard-snowleopard.merged-1.sucatalog"
cp
"/srv/SUS/html/content/catalogs/others/index-mountainlion-lion-snowleopard-
leopard.merged-1_$2.sucatalog"
"/srv/SUS/html/index-mountainlion-lion-snowleopard-leopard.merged-1.sucatal
og"
cp
"/srv/SUS/html/content/catalogs/others/index-10.9-mountainlion-lion-snowleo
pard-leopard.merged-1_$2.sucatalog"
"/srv/SUS/html/index-10.9-mountainlion-lion-snowleopard-leopard.merged-1.su
catalog"
cp "/srv/SUS/html/content/catalogs/index_$2.sucatalog"
"/srv/SUS/html/index.sucatalog"
;;

There is a command line option for this. This would be useful in the GUI.

Hey Guys,

Not sure if anyone else is running the netsus appliance on CentOS 7, but I ran into a few issues with the installation script for NetSUS_3.0.1.run. Just figured I should share some work arounds:

1. Install Apache, Samba, and Xinet prior to running the installation script.
2. You'll need to install these two dependencies (compat-db-headers-4.7.25-27.el7.noarch.rpm and compat-db47-4.7.25-27.el7.x86_64.rpm)
3. Once you kick off the install script, any other blanks should be filled.

NOTE: CentOS 7 no longer keeps daemons for httpd, smbd, and xinetd in /etc/ so the script will freak out and claim it's unable to start these services. You'll have to use "systemctl" to restart these services after the installation is complete.

I would advise creating a script you could execute in the future if you ever need to restart or possible upgrade NetSUS in the future. I'm still testing things out, but did notice "Could not save settingsCould not save settingsCould not save settings" always appears at the top of each webpage as you work your way through the web console for NetSUS. Not sure how much of a concern this is but will be sure to add more notes going forward if I stumble across anything new.

I had to change it from an .ovf to an .ova manually after the download to get VMWare ESXi 5 to be able to import it. Not sure if others are seeing the issue, but I thought I'd throw it out there.

The user guide outlines that it should be possible for users to access an SMB share on the NetSUS server.

Under Linux, a samba server requires smbd. In Ubuntu, this requires that the package samba be installed.

When the current 3.0 standalone installer (the coppy of the installer linked to from the front page) is run under Ubuntu LTS 12.04 server, however, samba does not appear to be installed as expected.

Steps to Reproduce

1. Start with a machine running a clean installation of Ubuntu LTS 12.04 Server. Make sure all of the packages are updated with apt-get.
2. On the clean Ubuntu machine from step 1, run the NetSUS 3.0 standalone installer linked to from the front page of this project.
3. Attempt to locate samba. I did this by running $ dpkg-query -l samba. If samba is not installed, this command returns an error message that no packages matching samba were found; otherwise, it returns name, version number, description, and other information for the currently-installed version of the samba package. Another easy way to do this is to look at /etc/smb: on a clean system this folder should exist only if samba is installed.

After running the installer, it is not possible to locate samba using either of the methods outlined in step 3.

Notes

I'm only reproducing this in Ubuntu 12.04. I'm not reproducing it in Ubuntu LTS 10.04 Server.

I cannot test this in an installer from a copy of NetSUS cloned from this repository (last checkin 152370d) due to #28.

My test machines are VirtualBox VMs, on a host running under OS X 10.9.2.

okay, so I got NetSUS installed on a fresh install of Ubuntu.

However when following the instructions to go to: https://myhostname.local/webadmin to access the NetBoot/SUS server web application, all any browser gives me is a combination of the following:

-Unable to connect
-can't establish a connection
-webpage not available
-connection refused

I've tried connecting via IP addresss, hostname.local, localhost, 127.0.0.1...and I get nothing.

What else do I need to do?
What on Earth am I doing wrong?

This is my first dive into Linux, so yes..I'm a bit clueless.

Hi Jamf,

Just running the installer script alone does not result in an installation on a basic red hat instance.. After a whole night of testing I found that the issue lays right at the beginning of the script.

Your "if and elif" statements to set ubuntuVersion based on lsb_release is faulty, as redhat also has a lsb_release (at least 6.4 and later do which I am using), therefore anything after that potentially fails as the process is assuming that I am installing on ubuntu and not red hat.

Hope this can be fixed in 3.1 and soon, as I have some new macs coming in next week with 10.9 and need a working SUS to be able to point them too.

My workaround was to rename lsb_release to lsb_release.orig then install, installation was ok, except for some reason it doesn't use https but defaults to http (not a prblem) the other issue is that the index.php and webadmin were saved under /var/www instead of /var/www/html which is the default, so i had to copy index.php and webadmin to /var/www/html.

After installation, I renamed lsb_release back and changed the line in adminHelper.sh in the if statement to look for lsb_releaseXXX just to get it to skip that part and move on to the next part which tells it this is a redhat/centos OS.

Other issues I am seeing it not adding an alias in httpd.conf for /srv/SUS is this normal? there is an alias for /srv/NETBOOT though.

Hope this helps.

Kamal

If NetSUS standalone is installed under a clean installation of CentOS 6.5 without first installing dependencies for netatalk, I'm observing output that appears to indicate both that the dependencies are not being installed as expected and that netatalk fails to install as expected.

I'm observing some output in the command line that appears to indicate that dependencies for netatalk are not being satisfied:

/usr/bin/perl is needed by netatalk-4:2.2.0-2.el6.x86_64
libavahi-client.so.3()(64bit) is needed by netatalk-4:2.2.0-2.el6.x86_64
libavahi-common.so.3()(64bit) is needed by netatalk-4:2.2.0-2.el6.x86_64
perl(File::Basename) is needed by netatalk-4:2.2.0-2.el6.x86_64
perl(File::Spec) is needed by netatalk-4:2.2.0-2.el6.x86_64
perl(IO::Socket) is needed by netatalk-4:2.2.0-2.el6.x86_64
perl(Socket) is needed by netatalk-4:2.2.0-2.el6.x86_64
perl(bigint) is needed by netatalk-4:2.2.0-2.el6.x86_64
perl(strict) is needed by netatalk-4:2.2.0-2.el6.x86_64
perl(vars) is needed by netatalk-4:2.2.0-2.el6.x86_64

In the very last line of output from the installer script, I'm also seeing an error message that appears to indicate that it was not possible to run netatalk (because the command does not exist at the path specified in the script):

NetSUSInstaller.sh: line 186: /etc/init.d/netatalk: No such file or directory

Steps to Reproduce

1. Run a clean installation of CentOS 6.5. Make sure all available updates have been applied with yum.
2. Per the instructions in the User Guide, install NetSUS from the standalone installer.
3. Watch the output returned by the installer for the error messages I outline above.

Notes

Netatalk is required to serve AFP from Linux, so to me (unless I'm missing something) it looks like this behavior likely breaks AFP access on a NetSUS server (where all of the STR are satisfied). I'm not familiar with NetBoot at all, but from my experience with netatalk (and some very-cursory research I did on DIY NetBoot servers in Linux), it seems like it might be worth understanding as well whether this could potentially affect netboot. If anybody with more experience in this area has any input, it seems like it might be helpful.

I was able to find some instructions to install netatalk 2.2.0 (the release it looks like NetSUS normally uses) here. If I install all dependencies other than rpm-build (which didn't seem relevant) between steps 1 and 2 in the STR I've provided, I'm able to install NetSUS successfully with netatalk (and I'm able to verify that netatalk is running as expected).

I'm not sure what the full dependency tree is for netatalk, but here is the command I run to make it work:

# yum install {openssl,pam,db4,libgcrypt,avahi,libacl,openldap,cracklib,quota,tcp_wrappers}-devel libtool avahi

I did not observe this at all under Ubuntu (LTS Server versions 12.04 and 10.04). In Ubuntu LTS Server 12.04 I observe that netatalk is not being installed (reported as #32). Under Ubuntu LTS Server 10.04, I'm observing that netatalk is installing as expected.

I do not currently have a copy of RHEL 6.4 (or later) that I can test this with. If at all possible, I recommend testing for this behavior under RHEL, as well.

It might also be worth noting that I reproduced this in a VirtualBox (v 4.3.10 r93012) virtual machine, with a host machine running under OS X 10.9.1.

If I build an installer from a clone of the current trunk/ main branch (last commit 152370d) and run the installer under Ubuntu LTS 12.04 server, the installer dies, posting the following error messages:

Checking for a supported OS...Error: Did not detect a valid Ubuntu/RedHat/Cent OS install.

Aborting installation due to unsatisfied requirements.

Steps to Reproduce

1. Run a copy of Ubuntu LTS 12.04 server where all available packages have been updated by apt-get.
2. Clone a copy of the current trunk/ main branch of NetSUS and build an installer from the source. To do this I used the copy of makeself included with the NetSUS project.
3. On the machine running Ubuntu 12.04 (the machine from step 1), run the installer built in step 2. When I did this I ran the installer as a non-root user with sudo.

When I do this, the installer fails after I am prompted for my sudo password (before I am prompted to confirm installing standalone).

Notes

As I note in the STR, the copy of makeself I used to build the installer that was included with this project. I built the installer with a copy of OS X 10.9.1. Xcode is not installed on the system the installer was built with.

If it helps at all, here is the command I used to clone the project to my local machine: $ git clone https://github.com/jamf/NetSUS.git

The machine I reproduced this on was a VirtualBox VM with the host running OS X 10.9.1.

I'm not reproducing this behavior under either CentOS 6.5 or Ubuntu LTS 10.04 Server. I'm only reproducing it under Ubuntu LTS 12.04 server.

I was not able to test this with RHEL 6.4 (or later).

I will post copies of /var/appliance/logs/applianceinstaller.log and the full script output in comments, below.

When I press the Enable NetBoot button on /webadmin/netBoot.php firebug tells me it is POSTing to /webadmin/netBoot.php with params:

NetBootImage CasperNetboot1093.nbi
enablenetboot Enable NetBoot

Unfortunately the netBoot.php script didn't appear to have any capability to actually handle this request so nothing happened. From looking over the code it appears that the enable netboot code was in ajax.php but was only set to trigger from changing the NetBoot Image dropdown select, not the Enable NetBoot button. I was able to get the Button to work by copying this code from ajax.php to netBoot.php and modifying the condition:

if (isset($_POST['NetBootImage']) && $_POST['enablenetboot'] = "Enable NetBoot")
{
$wasrunning = getNetBootStatus();
$nbi = $_POST['NetBootImage'];
if ($nbi != "")
{
$nbconf = file_get_contents("/var/appliance/conf/dhcpd.conf");
$nbsubnets = "";
foreach($conf->getSubnets() as $key => $value)
{
$nbsubnets .= "subnet ".$value['subnet']." netmask ".$value['netmask']." {\n\tallow unknown-clients;\n}\n\n";
}
$nbconf = str_replace("##SUBNETS##", $nbsubnets, $nbconf);
suExec("touchconf "/var/appliance/conf/dhcpd.conf.new"");
if(file_put_contents("/var/appliance/conf/dhcpd.conf.new", $nbconf) === FALSE)
{
echo "<div class="errorMessage">ERROR: Unable to update dhcpd.conf";

            }       
            suExec("disablenetboot");
            suExec("installdhcpdconf");

            if ($wasrunning || isset($_POST['enablenetboot']))
            {       
                    suExec("setnbimages ".$nbi);
            }       
            $conf->setSetting("netbootimage", $nbi);
    }       

}

With this change, pressing Enable NetBoot button worked again.

Josh

Edited out as I tried to install this over the top of our JDS which may cause issues in and of itself. I will test again on a true virgin CentOS 6.5 box.

So it happened. The VM drive I have set up is completely full with updates and I can no longer get into the web portal. I've tried SSH'ing into the VM and running the purge on deprecated files, but that has not created any space.

repoutil --purge-product all-deprecated --force

So I am at a loss. One question I do have is that currently I have 3 branches (beta, dev, and prod). If I have multiple branches does that mean that it downloads 3 copies of all the files? One for each branch.

The VM HD is set to 300GB like it was suggested. I am unsure how it filled up so fast.

Any help would be appreciated!

Have we given any thought to adding the feature of restricting the Netboot to only be available to certain MAC addresses, like we can do in Apple's Netinstall service? This could be handy in testing environments where multiple NetSUS's are set up and advertising.

It seems the NetBoot does not work properly on CentOS or RedHat. When trying to boot to the advertised NetBoot image a flashing globe appears on the client screen, after which it just boots to the hard disk. I am able to get it working on Ubuntu with the same image, the only difference being a spinning globe instead of a flashing one and it boots to the NetBoot image successfully. The /etc/dhcpd.conf file is exactly the same across OS's and all necessary files have the right privileges.

A unique log message I saw on the CentOS 7 server (/var/log/messages | grep dhcpd)

dhcpd: DHCPACK to <client-ip> <mac-address> via eno167777736
dhcpd: DHCPINFORM from <client-ip> via eno167777736
dhcpd: <company-name>-DHCP change
dhcpd: N Key or encapsulated
dhcpd: 68

I got nowhere trying to find the cause of this log message and I can't seem to find the cause of the issue anywhere in the NetSUSLP code.

There is currently a problem with the configuration of NetSUS 3.0. It may have been this way before, but I'm not sure. The problem is that the Apache configuration file is using redirect rules to redirect '/index.sucatalog' to the OS specific file in '/content/catalogs/others'. This file is not a copy of the 'Root Branch', it is a processed version of the .apple file, without any of the updates being excluded. The implication is that any client machine checking in to the root branch will have all updates enabled, even if specific updates are not enabled in the branch marked as the root branch.

Per the script which copies the root .sucatalog files, /var/appliance/sus_sync.py, the Apache config file should reference the .sucatalog files at the root of the share. This can be remedied by running the following two commands:

sed -i'.bak' -e 's|content/catalogs/others/||g' /etc/httpd/conf/httpd.conf
service httpd restart

If I install NetSUS under a clean installation of Ubuntu LTS Server 12.04, then reboot the machine, I observe upon reboot that a loop occurs where /var/appliance/dialog.sh posts (and repeats) an error message (that the command "dialog" was not found):

/var/appliances/dialog.sh: 21: /var/appliances/dialog.sh: dialog: not found

When this happens, I'm observing that the system is still responsive (and that I can connect to the system via SSH and issue commands). When I evaluate currently running processes with top, though, I'm observing that dialog.sh occupies 99.9% of the system CPU.

If I'm correct, this appears to occur because dialog is not installed by default in Ubuntu LTS Server 12.04. See my notes (below) for more information.

Steps to Reproduce

1. Run a clean installation of Ubuntu LTS Server 12.04. Make sure all available updates have been applied with aptitude.
2. Per the instructions in the User Guide, install NetSUS from the standalone installer.
3. Reboot the system. Once the system reboots (while the OS is loading), note the output that displays on the screen.

Notes

I'm only observing this under Ubuntu LTS Server 12.04. I'm not observing it at all under LTS Server 10.04, and I'm not reproducing it under CentOS 6.5. I do not currently have a copy of RHEL 6.4 (or later) that I can test this with.

If I install the package dialog (# apt-get install dialog) between steps 1 and 2 in my steps to reproduce, I'm able to follow the rest of the STR without reproducing this behavior.

In the interest of full disclosure (and regardless of whether it seems related), there were three things in my runtime environment that were exceptions to a 100% clean installation:

• I am using static IP (and not DHCP) in my network configuration.
• I have the ssh package installed (and sshd running).
• When installing the system I elected not to use LVM (I used a traditional HD configuration).

It also might be worth noting that I reproduced this in a VirtualBox (v 4.3.10 r93012) virtual machine, with a host machine running under OS X 10.9.1.

When i download a backup using the URL: https://NETSUS/webadmin/backup.php i cannot extract the backup.tar.gz file. I downloaded the file using Safari, Firefox and Chrome and tried to extract it using the terminal (tar -zxvf), stuffit expander, The unarchiver, GUI Tar.

After following the bonding process from https://help.ubuntu.com/community/UbuntuBonding

The Web console and smb sharing work fine but the NetBoot service will not start.

Could this be added to the network settings or a workaround be found

I installed Ubuntu on a PC and I can't get the netboot service to run.

The indicator is red always. Would someone please help me figure this out?

Harry

When I try to upload a NetBoot Image, I put in the username and password for smbuser and Finder hangs when connecting to server.

When installing CentOS 7.1 killall is not installed by default. You need killall to install it in order to Disable and Enable NetBoot.

Please install killall (part of psmisc) when installing running the NetSUS installer on CentOS.

Good Morning GitHub --

We have a customer that is looking for the ETA of the new OVA for JAMF Software's NetSUS. Please let us know if there an approximate date we can give them.

This is in reference to case: #21

Happy Monday!
Lois Fredrickson
JAMF Software

If NetSUS standalone is installed under a clean installation of Ubuntu LTS Server 12.04, apache/ httpd is not installed at the time NetSUS is. There are four things I'm seeing that seem to support this conclusion:

• If I attempt to access WebAdmin after installing (or after installing then rebooting), Chrome and cURL return messages that they were not able to access the server (cURL tells me the connection was refused).
• If I attempt to start apache/ httpd from the command line (# /etc/init.d/apache2 start) after running the installer, the command line returns an error message that it could not find the command issued.
• If I attempt to locate either apache or httpd with $ pgrep -l <keyword>, pgrep returns no results.
• The folder /etc/apache2 does not exist after the installer has run. In Ubuntu this is the path where binary- and config files for apache/ httpd are stored.

Steps to Reproduce

1. Run a clean installation of Ubuntu LTS Server 12.04. Make sure all available updates have been applied with aptitude.
2. Per the instructions in the User Guide, install NetSUS from the standalone installer.
3. Attempt to connect to WebAdmin. The server refuses the connection.
4. Try any of the other methods listed above.

Notes

When I tested this, I addressed #22 between steps 1 and 2. So I ran a clean installation of Ubuntu, addressed #22, and proceeded to run the installer.

I'm only reproducing this under version 12.04. I'm not able to reproduce this issue under Ubuntu LTS Server 10.04. I'm also not reproducing this under CentOS. I do not currently have a copy of RHEL 6.4 (or later) that I can test this with.

At one point earlier, I thought I was able to install successfully (and access WebAdmin) under 12.04 without reproducing this behavior. This was before I started keeping records of my testing, though, so I might be remembering incorrectly. If somebody else would be willing to take a look at this, though, before assigning to the developer, we should be able to determine whether it seems like what I remember here seems correct.

It might also be worth noting that I reproduced this in a VirtualBox (v 4.3.10 r93012) virtual machine, with a host machine running under OS X 10.9.1.

After running the NetSUS 4.1 installer on RHEL 6.7, I can log into the web GUI. However, I cannot change any settings such as any of the user passwords or any of the date and time settings. The web GUI seems to indicate that saving these settings was successful, but none of the settings are retained. Made sure SELinux was disabled, tested various wide open permissions, no change.

There is a command line way to do this already. Please enable it in the gui.

Add the following to you sus_sync.py

    os.system("cp /srv/SUS/html/content/catalogs/others/index-10.9-mountainlion-lion-snowleopard-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-10.9-mountainlion-lion-snowleopard-leopard.merged-1.sucatalog")

Here:

def sync_sus():
    print "Syncing SUS"
    os.system("/var/lib/reposado/repo_sync")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-lion-snowleopard-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-lion-snowleopard-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-leopard-snowleopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-leopard-snowleopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-mountainlion-lion-snowleopard-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-mountainlion-lion-snowleopard-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-10.9-mountainlion-lion-snowleopard-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-10.9-mountainlion-lion-snowleopard-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/index_" + strRootBranch + ".sucatalog /srv/SUS/html/index.sucatalog")

def enable_all_sus():
    print "Enabling new updates"
    for inst in instarr:
        print "Adding all updates to: " + inst
        os.system("/var/lib/reposado/repoutil --add-product all " + inst)
    os.system("cp /srv/SUS/html/content/catalogs/others/index-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-lion-snowleopard-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-lion-snowleopard-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-leopard-snowleopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-leopard-snowleopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-mountainlion-lion-snowleopard-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-mountainlion-lion-snowleopard-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/others/index-10.9-mountainlion-lion-snowleopard-leopard.merged-1_" + strRootBranch + ".sucatalog /srv/SUS/html/index-10.9-mountainlion-lion-snowleopard-leopard.merged-1.sucatalog")
    os.system("cp /srv/SUS/html/content/catalogs/index_" + strRootBranch + ".sucatalog /srv/SUS/html/index.sucatalog")

Newly created 14.04.1 Ubuntu server from linux.iso, only configured a couple of network settings (/etc/hosts, /etc/hostname, /etc/network/interfaces) restart then curled down the .run file... I see this error at the beginning.

NetSUSInstaller.sh: line 107: sestatus: command not found

Don't know if there needs to be a line to validate prior to 107 to see if sestatus is available. Ubuntu suggested to run "sudo apt-get install policycoreutils" to resolve.

I hacked this in your adminHelper.sh file, but there is a proper way to do it from within the sync command. This is needed to debug and verify syncing is working.

Please change from the current implementation of NetBoot to support BSDPy.

Making this change would allow the NetSUS to host multiple NBIs as well as display the real name of the NBI instead of "Faux NetBoot"

Currently NetSUS (3.0.2) supports dmg's but not sparse images without editing the adminHelper.sh file (https://derflounder.wordpress.com/2012/02/07/modifying-jamfs-netsus-to-host-deploystudio-netboot-boot-sets/).

I had an enhancement logged for AutoCasperNBI (macmule/AutoCasperNBI#4) to use sparse images, as this reduces NBI size from 8GB to 5GB (for example).

It should be feasible by changing the line in adminHelper.sh that contains: dmgfile=echo $files | grep dmg to dmgfile=echo $files | egrep dmg\|sparseimage as per what Rich has pointed out.

the /etc/dhcpd.conf file is doing weird things and the values inside it are not updating accordingly.

You can manually edit the values to make it work but once you go to Webadmin and press Enable Netboot it reverts back to those weird values.

I have tested this on a fresh install and an upgrade on Ubuntu 14.04.
This was happening on the appliance as well.

Please consider adding the version of the appliance into the GUI somewhere, perhaps within the About section.

Hello I ran the netsus installer but now I don't want it anymore, is there any command to purge it ?

Is this project still being actively maintained?

Many people continue to see problems or are limited in what they can do while there are open pull requests that resolve many of the issues and expand the usefulness of this tool.

Are the changes being audited by a code review and/or QA testing? If so, a simple comment would go a long way.