ESX deployment appliance v1.05

[brugh]

https://solutionexchange.vmware.com/store/products/esx-deployment-appliance-eda

ESX Deployment Appliance (EDA) is a small and easy to use appliance

that makes deploying ESX servers a breeze. It has a very intuitive

web-interface that can configure and deploy dozens of ESX servers in

minutes. It has a script-builder that will allow any admin to create

%post-scripts that will do most anything one needs to get the ESX hosts

up and running! Even if deploying with RDP/Altiris or the UDA, this

script-builder can help setting those up very quickly.

[brugh]

i got my own build working with esx5. i'll upgrade the appliance soon. just needs some scriplets and testing which usually takes a while..

[Arehnmark]

Any more news about how its going with the esxi 5 support?

[TariqMC]

Hi

I am trying to create 220 Hosts but the repeat process only works for 16 hosts starting at 1. If I create a second batch using the repeat process then the  "Starting count" seems to be ignored in in the IP address and it counts from 1 again.

Is there a file I can directly enter these into as I have all the information I need in a spreadsheet table?

Thanks

Tariq

[radus2k]

we looking for the latest one. Any clue about release date?

Please, please

[brugh]

i plan to do some final checks next tuesday so it should become publicly available that day or perhaps the next. sorry to keep you waiting, my internal build's been ready for a while. just need to 'applianice' it and do some checks with SD installations. vmware seems to actually support that with 5

[brugh]

and there it is! version 1.00 of the ESX Deployment Appliance. full support for all versions up to ESXi 5 and a scriptbuilder with a some new scriptlets specifically for ESX5 and some problems that a lot of people reported having problems with (like enabling management traffic on vmk adapters).

check it out on http://virtuall.nl/download-document/eda-v1.00.zip

please report any problems here. it's the only thread i regularly check and answer. any scriptlet ideas are very welcome too. i'll put the good ones in small updates that can quickly and easily put back into the EDA for everybody's benefit. enjoy!

[TariqMC]

Hi Brugh

One feature that I would love to see is the ability to team eda nics to get even better throughput and build 40-50

Esx servers in one go. At the moment I find that 25 is the max I can do in one go

Thanks

Tariq

Sent from my iPhone

[brugh]

Teaming in Linux isn't too hard but as most people dont deploy 50+ hosts at the same time i may not put that high on the request list

In case of the EDA though, you could try to add an extra NIC, then 'apt-get install ifenslave' and then change the /etc/network/interfaces to:

auto lo

iface lo inet loopback

auto bond0

iface bond0 inet static

  bond-slaves none

  bond-mode balance-rr

  address 172.16.1.99

  netmask 255.255.0.0

  gateway 172.16.1.254

auto eth0

iface eth0 inet manual

  bond-master bond0

  bond-primary eth0 eth1

auto eth1

iface eth1 inet manual

  bond-master bond0

  bond-primary eth0 eth1

and with '/etc/init.d/networking restart' you should have a loadbalancing bond0 adapter. this should work pretty well but you may want to add more memory to the eda first. could well be that memory runs out way before network bandwidth does .

also, i tweaked the /etc/apache2/apache.conf to not start too many servers at once. although maxclients is 150, you may want to change 'StartServers', 'MinSpaceServers' and 'MaxSpareServers' to values in the 5 to 10 range.

[vkadiyala]

what is the password to access dl$ share.

I tried the root password but it comes with access denied

[hughw23]

Hi Brugh,

I have just installed eda v1.0 and seem to be having a few issues.

The appliance takes a long time to boot > 20 mins, checked all VM performance which is fine, I the vm installed on an NFS share which has some disk latency, but this shouldn't be causing it to take this long to start up.

Second issue is on the EDA webpage, when I try to add some custom scripts using the drop down menu and click Add,

it adds them into the script which is all good, but when I click "save" it reverts back to the default script, have you experienced this before?

I experienced this in version 0.95 as well.

Thanks in advance.

[brugh]

the default password is root/root. if you change the root password on the config page, it should change the password for the dl$ accordingly.

[brugh]

the way you describe it, it sounds more like an external issue than an issue with eda itself.

what environment are you running eda in? does it have enough space? did you set limits on it somewhere?

[Zsoldier]

I'm running into a weird issue on a couple of my builds.  I have my ks script changing the local datastore name to match the hostname.  It's worked on several different hosts, but for some reason some hosts give me problems.  I also make changes to the local syslog path to match the renamed datastore, but for some reason I've noticed it filled in the Syslog.Local.DatastorePath variable w/ .locker/messages rather than the one I've specified.

I look @ the generated script on EDA and it looks fine, but the resulting ESX build results in the weirdness described above.  The hosts specified showing weirdness are listed 8th, 9th, and 10th out of a total of 21 entries listed.  Later entries haven't exhibited issues.

running 1.00.

Any ideas?

[brugh]

could you post the relevant part of the script and give a little detail on your setup? (dedicated deployment lan, host type, appliance environment perhaps?)

i can't imagine why nr 8 through 10 would fail and subsequent numbers would not. 

[Zsoldier]

It's really weird.  I think I figured out why the domain join isn't occurring, BIO's time and account cannot overwrite an existing account.

EDA resides on ESX Mgmt network.  It runs as DHCP server on this network.  Dell R710, Dell R810, Dell R900.  vSphere 4.1 Enterprise Plus.

To troubleshoot, I deleted the entries, rebooted the appliance, then started entering them again.  Still end up w/ wonky issues, but not on all hosts.  The only real relation is that the hosts I was having issues with are all part of the same cluster.  I've commented out below what seems to always work, vs. what seems to have issues:

# Bind vmnic4 to vSwitch0 (Always works)
# Active/Passive
esxcfg-vswitch -L vmnic4 vSwitch0
esxcfg-vswitch -X 1 vSwitch0

# Add vSwitch1, VMkernel, and bind vmnic1/3 (Always works)
# Active/Passive
esxcfg-vswitch -a vSwitch1
esxcfg-vswitch -A VMkernel vSwitch1
esxcfg-vswitch -L vmnic1 vSwitch1
esxcfg-vswitch -L vmnic3 vSwitch1
esxcfg-vswitch -X 1 vSwitch1
# Add vSwitch2 and bind vmnic2/5 (Always works)
# Active/Passive
# Use powershell script to add appropriate VM Networks
esxcfg-vswitch -a vSwitch2
esxcfg-vswitch -L vmnic2 vSwitch2
esxcfg-vswitch -L vmnic5 vSwitch2
esxcfg-vswitch -X 1 vSwitch2
# NTP Configuration (Always works)
/bin/echo "restrict 127.0.0.1" > /etc/ntp.conf
/bin/echo "restrict default kod nomodify notrap" >> /etc/ntp.conf
/bin/echo "server someserver.domain.com" >> /etc/ntp.conf
/bin/echo "server someserver1.domain.com" >> /etc/ntp.conf
/etc/init.d/ntpd restart
chkconfig --level 345 ntpd on
# Configures vmk0 as Mgmt Interface (Always works)
HOSTSVC_FILE=/etc/vmware/hostd/hostsvc.xml
/bin/cat > ${HOSTSVC_FILE} << __CREATE_HOST_SVC__
<ConfigRoot>
  <mangementVnics>
    <nic id="0000">vmk0</nic>
  </mangementVnics>
  <mode>normal</mode>
  <service>
    <tsm-ssh>off</tsm-ssh>
  </service>
</ConfigRoot>
__CREATE_HOST_SVC__
# Configure vmk1 as vMotion Interface (Problem with this.)
esxcfg-vmknic -a VMkernel -i ###HOSTIP### -n 255.255.255.0
/bin/vim-cmd hostsvc/net/refresh
/bin/vim-cmd hostsvc/vmotion/vnic_set vmk1

# DNS Secondary (Always works)
/bin/echo "nameserver 192.168.123.25" >> /etc/resolv.conf

# Configure Local Datastore Name to standard (Problems with this)
/bin/vim-cmd hostsvc/datastore/rename datastore1 "$(hostname -s)_local"

# Configure Syslog to Splunk (Problems with this)
/bin/vim-cmd hostsvc/advopt/update Syslog.Remote.Hostname string 10.64.216.18
/bin/vim-cmd hostsvc/advopt/update Syslog.Remote.Port int 514
# Only needed if local syslog going to be used.  (This seems to get filled w/ random info)
/bin/vim-cmd hostsvc/advopt/update Syslog.Local.DatastorePath string "[$(hostname -s)_local]/logfiles/$(hostname -s).log"

# Maps standard NFS Shares (Always works)
/sbin/esxcfg-nas -a -o server1 -s /share/DATA nfs_share
/sbin/esxcfg-nas -a -o server2 -s esx_vmtemplates nfs_share_ddns1564_vmtemplates

[brugh]

the script itself looks alright. it could be a timing issue though. the renaming may (for some odd reason) take some time (msec) and if you set the syslog location before the datastore is renamed the command may fail. you could try waiting a bit or move the store rename all the way to the top of the script. also, i'd use a variable to put the hostname in instead of generating it every time you need it. shouldn't matter but makes it more consistent script-wise.

[Zsoldier]

Good point.  I'll give that a try.  Another hypothesis I have is that I sometimes find myself switching between Japanese IME and English.  Wondering if I may be causing my own issues where I'm inserting characters in UTF format rather than standard ANSI.  Will test that too.

Still doesn't quite explain why the remote syslog entries are not getting filled out though.

Message was edited by: Zsoldier

[wuytens20111014]

Hi,

i started using the eda today (version 1.00) it looks very promising.

But i have some issues to get it working.

look at my screenshots.

I really want to start using this magnificant tool. So please help me resolve my issues.

thx a lot

[Vampire]

Sehr geehrte/r Absender/-in,

Sie erreichen mich ab sofort unter der Adresse: frank.hennen@ivz.de.

Die von Ihnen verwendete E-Mail Adresse ist nicht mehr weiter gültig. Eingehende Mails werden nicht gelesen und nicht weitergeleitet.

Für dadurch entstehende etwaige Unannehmlichkeiten bitte ich um Entschuldigung. Bitte ändern Sie meine Kontaktdaten in Ihrem persönlichen Adressbuch entsprechend ab.

Mit freundlichen Grüßen

Frank Hennen

[brugh]

do you have some details on your setup (hardware iscsi adapters? network setup?)

the installer tries to install to the first drive it finds. you could try changing the order in the bios.