Work – Page 21

Windows Cluster Service (continued)

June 1, 2008August 16, 2014 Christian 1 Comment

Well, guess my “solution” didn’t work sooo good. Lemme tell you what’s happening. I successfully added the node to the cluster group, but I can’t get *any* resources online.

The node tries bringing it online, then shows a failure and immidiately moves them over to the next node. There the resource is being successfully moved online .. So again, I’m out of ideas ..

Already tried reinstalling the box, after that I could get the third node successfully into the cluster, without the “Advanced (minimum)” trick … *shrug* still ain’t bringing any resources online.

IBM RDAC and Windows Cluster Service

May 28, 2008June 21, 2013 Christian 1 Comment

Okay, so we received a brand new x3650 the other day entitled to replace one (or better two) of our NAS frontend servers. We installed Windows on it the other day (had to create a custom Windows Server 2003 CD first, since the default one doesn’t recognize the integrated ServeRAID), and we prepped the box during the week with the usual things.

On Monday I started installing the “IBM StorageManager RDAC” MultiPath driver (since the box got two single port PCIe FC-HBA’s) and figured I’d be nice if we had this. I asked a IBM Systems Engineer of one of our partners, which told me generally there wouldn’t be a problem with Microsoft Cluster Services (MSCS) and the IBM MPIO driver. Only requirement would be that I’d install the new storport.sys driver (version 5.2.3790.4021) first (as in Microsoft KB932755).

Now, yesterday I finished the zoning, did the mappings on the storage arrays and then figured the box should see the hard disks. So I started adding another node to our existing Microsoft Cluster.

Result: Zip (as in MSCS telling me not all nodes could see the quorum disk)

Reason: a combination of two things. First, said IBM Storage Manager RDAC. The first time I installed it, I forgot about the storage mappings, thus the box seeing zero disks. After uninstalling it, I was seeing 121 (that’s right, one hundred and twenty one) new devices.

Visible volumes previous to installing the RDAC driver

That is basically a result of the zoning I did for this particular device, which has *all* controllers present in a single SAN zone, thus the HBA’s seeing devices eight (or nine) times .. Update: yes, I’m missing one controller … 😀

Now, as I reinstalled the RDAC *after* the host discovered the volumes, it’s showing only a dozen drives.

Visible volumes after installing the RDAC driver

Now, as I figured this out, I told myself “Hey, adding the third node to the Windows Cluster should now work without a clue …” … guess what ?

It’s Microsoft and it doesn’t. Now why doesn’t it work ? ‘Cause the Cluster Setup Wizard is getting confused in Typical mode, as it’s creating a “local quorum disk” which naturally isn’t present in the cluster it’s joining. Now, switching the wizard to “Advanced (minimum) configuration” as suggested in Q331801, just works … *shrug*

Windows XP Embedded and GPO settings (continued)

May 26, 2008August 16, 2014 Christian 1 Comment

Well, as I said in my previous post, I do have some weird things happening. Apparently adding the domain user to the local group “Administrators” makes everything just works fine, yet he can’t do administrator like stuff (like turning off the write protection, changing local user accounts, …).

Also, if you’re looking for a smart way of how to add a certain global group (as in Active Directory group) to a local group, try this:

NET LOCALGROUP Administrators /ADD DOMAINGROUPNAME

1	NET LOCALGROUP Administrators /ADD DOMAINGROUPNAME

That simple, doesn’t even need the usual credentials to lookup the object, it apparently bypassed that step *shrug*.

And yet another weird thing is: if I run a certain command from a deployment script, it gives me different result as a manual execution of said script would give me .. *shrug*

; html-script: false  ]NETDOM JOIN %COMPUTERNAME% /domain:heimdaheim.de 
  /OU:"OU=Thinclients,OU=Computers,DC=heimdaheim,DC=de" 
  /UserD:%ADMIN% /PasswordD:somepass 
  /User0: Administrator /Password0:Administrator

NET LOCALGROUP Administrators /ADD BARFOODomain-Users

; html-script: false ]NETDOM JOIN %COMPUTERNAME% /domain:heimdaheim.de

/OU:"OU=Thinclients,OU=Computers,DC=heimdaheim,DC=de"

/UserD:%ADMIN% /PasswordD:somepass

/User0: Administrator /Password0:Administrator

NET LOCALGROUP Administrators /ADD BARFOODomain-Users

If I put that into a rsp (that is Wyse Device Manager script), it ain’t working. Would I be executing it myself without the WDM, everything works like a charm … *yuck*

Rescuing a rebooting machine that’s hanging

May 24, 2008June 21, 2013 Christian Leave a comment

One of my co-worker approached me today with a weird problem. Yesterday he had a disk in a 900GiB array failing which he replaced. After that, he run a rebuild/verification, fsck’ed the file system and tried to mount the volume again.

Apparently the mount produced a kernel oops (guess what, the 900GiB is running reiserfs), thus leaving the kernel tainted (or what do they call it ?). So he tried to reboot the box but it didn’t reboot. It started rebooting but then hung (as in not continuing the reboot). He tried to ssh back to the box, and it worked just fine.

This is where sysrq comes in handy.

# This is gonna activate the sysrq
echo 1 &gt; /proc/sys/kernel/sysrq

# Now, since we ain&#039;t at a console, we can&#039;t use the sysrq keys
# (&quot;b&quot; for reboot, &quot;o&quot; for shutdown)
echo b &gt; /proc/sysrq-trigger

# This is gonna activate the sysrq

echo 1 > /proc/sys/kernel/sysrq

# Now, since we ain't at a console, we can't use the sysrq keys

# ("b" for reboot, "o" for shutdown)

echo b > /proc/sysrq-trigger

That’ll restart the box, and cha-ching .. 😀

IBM (Tivoli) Integrated Solutions Console

May 23, 2008June 21, 2013 Christian Leave a comment

Here I am, preparing our environment for the first (of hopefully many) tester for our upcoming VTL project. So I ended up installing the ISC and Administration Center for Tivoli Storage Manager on a 64bit guest (that is SLES10 for AMD64), just because I forgot to include support for later versions with our current running one. Guess what, na–na na na na. Exactly, didn’t work, the same errors I got while trying it before in a virtual environment. “Portlet is not available.”

So I ended up redoing the whole thing on a 32bit guest and guess what … bada-bing. works … *shrug* I don’t know whether or not that’s a surprising thing .. but what surprises me, is that I do have a working 64bit Integrated Solutions Console and Administration Center running, only difference is that one is running on real hardware.

Anyway, after looking on how the Integrated Solutions Console (that is the Websphere environment – yes *yuck*) did it’s own start up after boot (you know, I’d like to restart an application if it’s hanging without the need to reboot the whole box), I found this particular line of code:

; html-script: false ]isc6:23:boot:/opt/tivoli/isc/PortalServer/bin/startISC.sh ISC_Portal

1	; html-script: false ]isc6:23:boot:/opt/tivoli/isc/PortalServer/bin/startISC.sh ISC_Portal

And since I was lazy (and it was already Friday afternoon), I ended up writing a small init script which rids you of the need of such a ugly way to start a service.

#!/bin/sh
#
# /etc/init.d/isc
#
### BEGIN INIT INFO
# Provides:       isc
# Required-Start: network
# Should-Start:
# Required-Stop:  network
# Default-Start:  2 3 5
# Default-Stop:
# Description:    Start the Tivoli Integrated Solutions console
### END INIT INFO

. /etc/sysconfig/isc

BINDIR=/sbin

. /etc/rc.status
rc_reset

case &quot;$1&quot; in
  start)
    echo -n &quot;Starting Tivoli Integrated Solutions Console server&quot;
    startproc -p /var/run/tivoli-isc.pid 
      $ISC_INSTALL/PortalServer/bin/startISC.sh 
      ISC_Portal 2&gt;&amp;1 &gt;/var/log/tivoli-isc.log
    # I know, this is *real* ugly
    PID=$( tail -n 1 /var/log/tivoli-isc.log | cut -d  -f10 )
    echo $PID &gt;/var/run/tivoli-isc.pid
    rc_status -v
    ;;
  stop)
    echo -n &quot;Stopping Tivoli Integrated Solutions Console server&quot;
    $ISC_INSTALL/PortalServer/bin/stopISC.sh 
      ISC_Portal $ISC_ADMIN $ISC_PASSWORD 
      2&gt;&amp;1 &gt;/var/log/tivoli-isc.log
    rc_status -v
    ;;
  restart)
    $0 stop
  	$0 start
  	rc_status
  	;;
  status)
    echo -n &quot;Tivoli Integrated Solutions Console server:&quot;
    checkproc -p /var/run/tivoli-isc.pid 
      $ISC_INSTALL/AppServer/java/bin/java
    rc_status -v
    ;;
  *)
  	echo &quot;Usage: $0 {start|stop|status|restart}&quot;
    exit 1
  ;;
esac
rc_exit

#!/bin/sh

# /etc/init.d/isc

### BEGIN INIT INFO

# Provides: isc

# Required-Start: network

# Should-Start:

# Required-Stop: network

# Default-Start: 2 3 5

# Default-Stop:

# Description: Start the Tivoli Integrated Solutions console

### END INIT INFO

. /etc/sysconfig/isc

BINDIR=/sbin

. /etc/rc.status

rc_reset

case "$1" in

start)

echo -n "Starting Tivoli Integrated Solutions Console server"

startproc -p /var/run/tivoli-isc.pid

$ISC_INSTALL/PortalServer/bin/startISC.sh

ISC_Portal 2>&1 >/var/log/tivoli-isc.log

# I know, this is *real* ugly

PID=$( tail -n 1 /var/log/tivoli-isc.log | cut -d -f10 )

echo $PID >/var/run/tivoli-isc.pid

rc_status -v

;;

stop)

echo -n "Stopping Tivoli Integrated Solutions Console server"

$ISC_INSTALL/PortalServer/bin/stopISC.sh

ISC_Portal $ISC_ADMIN $ISC_PASSWORD

2>&1 >/var/log/tivoli-isc.log

rc_status -v

;;

restart)

$0 stop

$0 start

rc_status

;;

status)

echo -n "Tivoli Integrated Solutions Console server:"

checkproc -p /var/run/tivoli-isc.pid

$ISC_INSTALL/AppServer/java/bin/java

rc_status -v

;;

echo "Usage: $0 {start|stop|status|restart}"

exit 1

;;

esac

rc_exit

And the corresponding sysconfig file:

## Type:     string
## Default:  iscadmin
## Config:   &quot;&quot;
## ServiceRestart:  isc
#
# Adminstration user for the Portal
ISC_ADMIN=&quot;iscadmin&quot;

## Type:     string
## Default:  iscadmin
## Config:   &quot;&quot;
## ServiceRestart:  isc
#
# Adminstration user password for the Portal
ISC_PASSWORD=&quot;iscadmin&quot;

## Type:     string
## Default:  /opt/ISC/601
## Config:   &quot;&quot;
## ServiceRestart:  isc
#
# Full path to the ISC Portal installation root directory
ISC_INSTALL=&quot;/opt/tivoli/isc&quot;

## Type: string

## Default: iscadmin

## Config: ""

## ServiceRestart: isc

# Adminstration user for the Portal

ISC_ADMIN="iscadmin"

## Type: string

## Default: iscadmin

## Config: ""

## ServiceRestart: isc

# Adminstration user password for the Portal

ISC_PASSWORD="iscadmin"

## Type: string

## Default: /opt/ISC/601

## Config: ""

## ServiceRestart: isc

# Full path to the ISC Portal installation root directory

ISC_INSTALL="/opt/tivoli/isc"

Et voilá, it’s done. Now just a `chkconfig -a isc‘ and it’s gonna startup nice and easy (when it really should) via the normal service startup and not get spawned from the inittab.

Windows XP Embedded and GPO settings

May 19, 2008June 21, 2013 Christian 3 Comments

We’re currently having a weird issue (which we had before); the Windows XP Embedded powering our Wyse V90’s isn’t applying any GPO settings if you log on with a user that has a configured profile.

Googling (is that a valid word yet ?!) for it, only resulted in one useful link, which is apparently a guy with the exact same problem … *shrug* I’m completely out of ideas by now, as I don’t even have a place to start (as in where the reason might be located).

Well, I do have a place to start with (that’s the local Events Viewer), which indeed lists some errors, but only such errors which ain’t making any sense. For example I see this:

Userenv:1086 – “Windows cannot do loopback processing for downlevel or local users. Loopback processing will be disabled.“
SceCli:1704 – “Security policy in the Group policy objects has been applied successfully.“
Userenv:1085 – “The Group Policy client-side extension Folder Redirection failed to execute. Please look for any errors reported earlier by that extension.“

Getting a FC HBA to rescan it’s attached devices

May 19, 2008June 21, 2013 Christian Leave a comment

If you’re using a 2.6 based distribution, the FC HBA (or more correctly the corresponding driver) should create entries in /sys/class/scsi_host. Now you only need to get the host-number (basically the # of the host bus adapter) and you can get started ..

Simply doing this, is going to tell the FC HBA “rescan” and discover new devices ..

echo &quot;1&quot; &gt; /sys/class/fc_host/host/issue_lip

1	echo "1" > /sys/class/fc_host/host/issue_lip

That should do the trick, and you should be able to get udev to recognize the new devices attached via FibreChannel without the need to reboot the whole box (which might be a bit tricky).

patch2mail for SLES10

April 21, 2008June 21, 2013 Christian 1 Comment

Well, there is this “nifty” tool called patch2mail, which basically converts the XML for the updates to a more readable format. But you’re screwed if you want to do the same on SLES10. Since it ain’t shipping with the zypper xml wrapper thing, you need to do it a bit different.

So I ended up writing a small (and yet, ugly) shell script to generate me a mail of my liking ..

#!/bin/bash
TO=&quot;admin-addr@localhost&quot;
CLASSES=&quot;security recommended optional&quot;
 
# Temporary files
ZYPP_LIST=&quot;$( mktemp /tmp/zypper-list.XXXXXX )&quot;
ZYPP_DETAILS=&quot;$( mktemp /tmp/zypper-details.XXXXXX )&quot;
TMP=&quot;$( mktemp /tmp/zypper-report.XXXXXX )&quot;
zypper pch 2&gt;/dev/null &gt; $ZYPP_LIST
 
# Figure out how much updates are still pending
PENDING=&quot;$( cat $ZYPP_LIST | grep &quot;| Needed&quot; | wc -l )&quot;
 
if [ $PENDING -eq 0 ] ; then
  exit 0
fi
 
echo &gt; $TMP
echo &quot; Pending updates for $( domainname -f ) on $( date )&quot; &gt;&gt; $TMP
 
for severity in $CLASSES; do
  PACKAGES=&quot;$( cat $ZYPP_LIST | egrep &quot;${severity}(.*)| Needed&quot; | 
    cut -d| -f2 | sed &quot;s,^ ,,&quot; )&quot;
  echo
  echo &quot;  Category: $severity&quot;
  for package in $PACKAGES; do
    zypper patch-info $package 2&gt;/dev/null &gt; $ZYPP_DETAILS
    echo &quot;  * $package ($( cat $ZYPP_DETAILS | grep &quot;Version: &quot; | 
      sed &quot;s,Version,version,&quot; ))&quot;
    echo &quot;    $( cat $ZYPP_DETAILS | grep &quot;Summary: &quot;)&quot;
    echo &quot;    $( cat $ZYPP_DETAILS | grep &quot;Reboot Required:&quot; )&quot;
    echo
  done
  echo
done &gt;&gt; $TMP
 
cat $TMP | 
  mail -s &quot;[$( date +%F )] Update report for $( domainname -f )&quot; $TO
trap &#039;rm -f &quot;$TMP&quot; &quot;$ZYPP_LIST&quot; &quot;$ZYPP_DETAILS&quot; &gt;/dev/null 2&gt;&amp;1&#039; 0
trap &quot;exit 2&quot; 1 2 3 15
 
# vim: set tw=80

#!/bin/bash

TO="admin-addr@localhost"

CLASSES="security recommended optional"

# Temporary files

ZYPP_LIST="$( mktemp /tmp/zypper-list.XXXXXX )"

ZYPP_DETAILS="$( mktemp /tmp/zypper-details.XXXXXX )"

TMP="$( mktemp /tmp/zypper-report.XXXXXX )"

zypper pch 2>/dev/null > $ZYPP_LIST

# Figure out how much updates are still pending

PENDING="$( cat $ZYPP_LIST | grep "| Needed" | wc -l )"

if [ $PENDING -eq 0 ] ; then

exit 0

echo > $TMP

echo " Pending updates for $( domainname -f ) on $( date )" >> $TMP

for severity in $CLASSES; do

PACKAGES="$( cat $ZYPP_LIST | egrep "${severity}(.*)| Needed" |

cut -d| -f2 | sed "s,^ ,," )"

echo

echo " Category: $severity"

for package in $PACKAGES; do

zypper patch-info $package 2>/dev/null > $ZYPP_DETAILS

echo " * $package ($( cat $ZYPP_DETAILS | grep "Version: " |

sed "s,Version,version," ))"

echo " $( cat $ZYPP_DETAILS | grep "Summary: ")"

echo " $( cat $ZYPP_DETAILS | grep "Reboot Required:" )"

echo

done

echo

done >> $TMP

cat $TMP |

mail -s "[$( date +%F )] Update report for $( domainname -f )" $TO

trap 'rm -f "$TMP" "$ZYPP_LIST" "$ZYPP_DETAILS" >/dev/null 2>&1' 0

trap "exit 2" 1 2 3 15

# vim: set tw=80

Software support and “key account” managers

April 14, 2008August 16, 2014 Christian Leave a comment

As Mike wrote about his experiences with hardware vendors, I’m gonna devote this here post to my favorite software company in the world. We recently bought two copies of a software called “2X Application Server Enterprise Edition“. As one would think from reading the spe cs of the software, it’s near a Citrix solution (which it is, at least for a small part); but in return it’s faaaar away concerning the price. Just so you get an idea, about what I’m meaning with “faaar“:

Windows Server 2003:
Standard Edition:        2 * 91,00
CAL:                    50 *  6,00
Terminal Server CAL:    50 * 17,00
                       ___________
                          1.332,00

Windows Server 2003:

Standard Edition: 2 * 91,00

CAL: 50 * 6,00

Terminal Server CAL: 50 * 17,00

___________

1.332,00

The above are fixed costs, you need them anyway as both Citrix as well as the 2X solution is only working *on top* of Windows Server 2003 Terminal Services.

Now, here’s the real comparison between 2X Application Server & Loadbalancer and Citrix XenApp Platinum Edition:

2X               2 * 1510,00 =  3.020,00
Citrix          50 *  393,00 = 19.650,00

1 2	2X 2 * 1510,00 = 3.020,00 Citrix 50 * 393,00 = 19.650,00

While 2X is licensed per terminal server, XenApp is licensed per user. As you can see from the above prices, the 2X solution is roughly 1/6 of the Citrix XenApp solution.