Available for linux is the sopcast command line player – sp-sc-auth which works just fine, so all I needed to go was get this to my ipad. Fairly straight forward I thought as sopcast listens on a chosen port for you to connect a media application to – I tried it on my ipad with xyzplayer which worked fine for a standard def video stream, but with HD streams the picture would be in slow motion.

I tried a number of other players, but they all had the same problem. So I figured if I could re-stream the stream as h264 I could get quicktime to play it on my ipad, and that would sort the issue.

However – easier said than done!

I was able to get ffserver and ffmpeg to take the stream from sopcast and re-stream but again slow motion issues.

So I definitely needed to re-stream into a quicktime compatable stream.

So I started looking at the use of the open source conversion of apples segmenter. Looked great but had a nightmare trying to get live_segmenter.c to compile on my fedora system.

However after some digging it seemed the majority of my issues were with my version of ffmpeg.

Using these instructions I managed to get it to compile correctly:

wget http://epirat.de/wp-content/uploads/ffmpeg-export-snapshot-2009-12-02.tar.bz2

tar -xvjf ffmpeg-export-snapshot-2009-12-02.tar.bz2

cd ffmpeg-export-2009-12-01/

./configure --enable-gpl --enable-nonfree  --enable-libfaac\
 --enable-libmp3lame --prefix=/tmp/old_ffmpeg

make

make install

cd

git clone git://github.com/ePirat/HTTP-Live-Video-Stream-Segmenter-and-Distributor.git

gcc -v -Wall -g live_segmenter.c -o live_segmenter \
    -lavformat -lavcodec -lavutil -lvorbis -ltheora\
    -lbz2 -lm -lz -lfaac -lmp3lame \
    -I/tmp/old_ffmpeg/include \
    -L/tmp/old_ffmpeg/lib

rm -rf /tmp/old_ffmpeg

rm -rf ffmpeg-export-2009-12-01

rm ffmpeg-export-snapshot-2009-12-02.tar.bz2

I would love to credit the author but im unable to find a way too! So if you read this thanks, and drop me a mail for credit

After getting this to compile, I followed the instructions http://www.ioncannon.net/programming/452/iphone-http-streaming-with-ffmpeg-and-an-open-source-segmenter/

Well couldnt get the script to work so stripped it out to manually run ffmpeg and pipe it into live_httpstreamer – however turns out my m3u8 index wasnt being created and my files were not able to be encoded as I didnt have x264 installed correctly.

So took this https://sites.google.com/site/linuxencoding/install-script to get the x264 support installed.

Manually created my own m3u8 index just to test that it would actually work.

Ended up with it all working, although the audio is going out of sync, but i’ll look at that at some point..

My final command, assuming I have created the m3u8 file with just 2 hours worth of segments, is…

/home/user/Linux.Encoding.install/bin/ffmpeg -i http://127.0.0.1:7070 -f mpegts -acodec libmp3lame -ar 32000 -ab 48k -async 1 -s 600×480 -vcodec libx264 -b 1200k -flags +loop+mv4 -cmp 256 -partitions +parti4x4+partp8x8+partb8x8 -subq 7 -trellis 1 -refs 5 -coder 0 -me_range 16 -keyint_min 25 -sc_threshold 40 -i_qfactor 0.71 -bt 1200k -maxrate 1200k -bufsize 110k -rc_eq ‘blurCplx^(1-qComp)’ -qcomp 0.6 -qmin 10 -qmax 51 -qdiff 4 -level 30 -aspect 4:3 -r 30 -g 90 -async 2 – | /home/user/segmenter/live_segmenter 10 /var/www/html/video/ stream ipad

Any questions please feel free to post comments, I have pulled this together quickly so might well have missed something…

Much googling turned up nothing however playing around I found that the daemon died immediately when I turned calendar off of notification centre and back to alerts rather than banner alert. I am hopeful that this will help the massive battery drain I have been experiencing and I will report back on this post.

Whilst it is nice to have my calendar stuff in the notification centre with the imminent release of intelliscreenX I can live without it for the moment if it gives me better battery life.

Cheers

I put it down to spotlight re-indexing etc and ignored it for a few days, but after a week of still being rubbish I did some digging.

Although not conclusive, I noticed a process – calaccessd which was using about 40% of the cpu time all the time – obviously cpu time equates to battery drain.

After some more digging I discovered its related to the calendaring functions (yes slightly obviously given the name). Doing some debugging, I went into my calendars and unticked the iCloud calendar from my views – this immediately made the calaccessd drop cpu use to next to nothing.

So im hopeful this might have a significant effect on my battery life and put it somewhere back to where I expect it to be..

Any comments on this i’d be interested to hear.

Cheers!

update1 – checked my running processes a few hours later and irritatingly calaccessd was back using lots of cpu, so this time I have switched calendar off completely from my iCloud mail settings – this so far has stopped any resurgence of calaccessd using lots of cpu.

update2 – this has fixed the cpu runaways, but still experiencing very poor battery life, and very jerky and slow performance from the phone as a whole. Going for a full restore to see how it fairs after that..

update3 – did the full restore, again had fun and games with stupid itunes not recognizing my backup, but battery life is now back to normal, if not better! I had never noticed but each time i pick up the phone now it actually feels cold, which must be a good thing as if its warm its obviously using battery up!

I did choose to not install a couple of items I had previously installed, and have removed a couple of calendars also so not conclusive but its got to be one of the following :

poweralerter, sbsettings, iCloud syncing calendars etc, multiple exchange calendars from the same account (I had 4 calendars on my 1 exchange account, which I have trimmed down to 1 now)

So its either 1 of the above or just the restore, but either way im considering this case closed. Its also worth noting that its performing lovely too, the jerkyness has gone and its nice and nippy. All in all im happy

update4 not happy again as per the above update I thought it was sorted, however it seems not. See post what is calaccessd for a further update.

update5 – post on calaccessd gives full info on the bug on ios 5, however updating to ios 5.0.1 fixes this issue.

After some googleing I came up with the following script with a commented out line that would allow a copy of those files after indexing.

———————————————————————————-

$path = Read-Host “Please enter the top-level path (eg: C:\Temp)”

$user = Read-Host “Please enter the user to be searched for (eg: DOMAIN\User)”
$dst = Read-Host “Please enter the copy destination:”

$files = Get-childitem $path -recurse |
where {$_.lastwritetime.date -gt
[datetime]::parse(“01/01/2010″) -and
$_.lastwritetime.date -lt
[datetime]::parse(“01/01/2011″)}

foreach ($file in $files){
$owner = Get-Acl $file.FullName
if ($owner.Owner -eq $user){Write-output $file.FullName >> output}
#if ($owner.Owner -eq $user){copy-item -path $file.FullName -dest $dst -force}
else {}
}

You might need to do this if you have an ip address conflict or for some other reason where going through all the guests would be a time consuming and laborious process.

With ESX and ESXi version 4 onwards that command was removed, so the below ash script will allow you to either list the VM’s on a host, or if you use it with an argument it will search for a mac address or part of a mac address.

This script will allow you to search and find a mac address and tie that to a guest name. The script needs to be run on the ESX or ESXi host/s

run it from the command line like below

#./macfinder.sh   < no arguments to list all registered VM’s

or

#./macfinder.sh 00:50:56:00:00:00  < with argument to find specific or part of mac address

___________________________ shell script below —————————

#!/bin/ash
# searches current registered vm’s then uses that list to find macAddresses of those

# then greps output for desired mac or part of mac address

vim-cmd vmsvc/getallvms | awk ‘{print $1″:” $2}’ | grep -v Vmid > /tmp/allvms
for lines in `cat /tmp/allvms`do

id=$(echo $lines | awk -F: ‘{print $1}’)

name=$(echo $lines | awk -F: ‘{print $2}’)

mac=$(vim-cmd vmsvc/device.getdevices $id | grep macAddress)

if test $1then
echo $name $mac | grep $1

else echo $name $mac

fi

done

rm -rf /tmp/allvms

The config below is what is required on the cli – if you have access to the web interface its pretty straight forward there too.

port-mirroring mode xrxorxtx monitor-port 1/23 mirror-port-X 2/3 mirror-port-Y 2/4

According to the nortel documentation this mirrors ports 2/3 and 2/4 to port 1/23 – however in practice and looking through the bugs list the firmware we ran does not actually support monitoring 2 ports, so we ended up falling back to just the single mirror.

But it is possible using the underlying system.

To create a route that allows all traffic from 10.0.0.0/24 to route through interface 5 try the following:

We add a new routing table by adding to the file:

echo "100 Defaultroute2" >> /etc/iproute2

This will create a new routing table called Defaultroute2.The default routing tables are numbered 253 & above, so as the number assigned to this table is less than 253, ie: 100, it will be used BEFORE the default routes. Next we set up a new rule for this table:

ip rule add from 10.0.0.0/24 table Defaultroute2

This adds a rule that says any traffic FROM 10.0.0.0/24 will have the routing in the Defaultroute2 routing table applied. Next we set up some new routes:

ip route add default via 192.168.0.1 dev Lan5 table Defaultroute2

This sets a default route that pushes traffic to the gateway address 192.168.0.1 via the device Lan5. This route applies to any traffic going via the Defaultroute2l routing table. We can add more routes to this table, so for instance we could route traffic to other IP addresses via a different gateway:

ip route add 212.111.123.0/24 via 212.104.137.254 dev Lan5 table Defaultroute2

and bingo you have not implemented source based routing.

Dont forget however that this config will need to be applied on each boot, or else create a simple startup script.

Anyway the fix is as follows:

login to the box and ensure you are in expert mode.

then do

cd $FWDIR

cd database

mv fwauth.NDB fwauth.corrupt

thats it!

Reinstall your firewall policy and you should now see it sucessfully install and recreate the fwauth.NDB database.

It took me some digging to find the cause, and it would seem that for some reason – I havent invested the time to find out why yet – that the table uprofiles_tasks gets so big that it causes problems to the login script and prevents logins.

The answer is to login to the mysql database, and truncate the uprofiles_tasks table.

This then fixes logins for a period of time until the table gets large again, but in my experience this is weeks / months rather than days..

The cause of the problem was a iscsi lock caused by 2 hosts trying to write to the same store at the same time.

This was evident from messages on the ESX hosts from the error message: vprob.vmfs.heartbeat.timedout and referencing one of the volumes on our ISCSI storage.

This was causing the entire ESX host to have connectivity problems as well as affecting the guests that resided on that volume.

Because the host and guests were not accesible through vSphere, we were unable to remove the volume or power cycle the guests.

After much digging around and with the help of VMWare support we understood that the cause of the problem was a lock on that filesystem, and to fix the problem we ran vmkfstools -L lunreset /vmfs/devices/disks/volumename…

This removed the lock which was caused by 2 hosts trying to write to the same volume at the same time and causing a iscsi lock.

Very painful, but happy to have the cluster back up and running.