Quantcast
Channel: StudioSysAdmins Message Board
Viewing all 3749 articles
Browse latest View live

Blackmagic UltraStudio 4K - HDMI output and splitters

$
0
0
Blackmagic UltraStudio 4K - HDMI output and splitters
posted by Jean-Francois Panisset on Jan. 7, 2016, 11 a.m.
Sounds like the wrong EDID info is being presented to the UltraStudio by the Key Digital splitter. According to:

http://keydigital.com/downloads/phantoms/KD-HD1x2_1x4ProK_Manual.pdf

Try setting the rotary switch to the resolution you want, either 1080/60i or 1080/60p.

EDID is always a pain...

JF


On Thu, Jan 7, 2016 at 7:51 AM, Thomas Ericsson <thomas.ericsson@fido.se> wrote:
Does anyone out there have a working combo that splits UltraStudio HDMI output to 2 screens. We currently run the following setup:

MacPro -> UltraStudio via Thunderbolt -> Splitter: Key Digital KD-HD1x2ProK via HDMI -> NEC SpectraView Reference 322UHD + Sony Bravia FW-55X8570C via HDMI.

As soon as we put the splitter in the mix the UltraStudio gets confused and outputs the wrong resolution for both devices. A Quantum video signal tester outputs the correct signal to both devices through the splitter, so I assume the problem is with the UltraStudio. BlackMagic is of no help, so we will check the forums there as well, but i thought i would try here as well.

Thomas

--
-------------------------------
Fido
Rosenlundsgatan 40, 4 tr.
118 53 Stockholm
SWEDEN
-------------------------------

To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe

Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
Sounds like the wrong EDID info is being presented to the UltraStudio by the Key Digital splitter. According to:

http://keydigital.com/downloads/phantoms/KD-HD1x2_1x4ProK_Manual.pdf

Try setting the rotary switch to the resolution you want, either 1080/60i or 1080/60p.

EDID is always a pain...

JF


On Thu, Jan 7, 2016 at 7:51 AM, Thomas Ericsson <thomas.ericsson@fido.se> wrote:
Does anyone out there have a working combo that splits UltraStudio HDMI output to 2 screens. We currently run the following setup:

MacPro -> UltraStudio via Thunderbolt -> Splitter: Key Digital KD-HD1x2ProK via HDMI -> NEC SpectraView Reference 322UHD + Sony Bravia FW-55X8570C via HDMI.

As soon as we put the splitter in the mix the UltraStudio gets confused and outputs the wrong resolution for both devices. A Quantum video signal tester outputs the correct signal to both devices through the splitter, so I assume the problem is with the UltraStudio. BlackMagic is of no help, so we will check the forums there as well, but i thought i would try here as well.

Thomas

--
-------------------------------
Fido
Rosenlundsgatan 40, 4 tr.
118 53 Stockholm
SWEDEN
-------------------------------

To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe


Rehat 5.11 Fileserver dual nic,dual subnet config

$
0
0
Rehat 5.11 Fileserver dual nic,dual subnet config
posted by Brian Smith on Jan. 8, 2016, 10:39 a.m.

I have two different switches that go to my sonicwall firewall, 

 

switch a is on the 10.0.1.x subnet and switch b is on the 10.0.2.x subnet.

 

On switch A is a Redhat 5.11 File server with static network IP 10.0.1.5 via eth2 which is a 10gbe etehrnet card.

 

right now the server  has 2 active network connections - 

IB0 is the infiniband connection to a separate subnet 192.168.100.x this network is local and doesn't goto the firewall, and has a bunch of servers and the other flames on it used for speedier file transfers for the flames.

 

eth2 which is a 10gbe nic and is currently connected to the switch on 10.0.1.x and gives me access to 

network, 10.0.1.5/255.255.255.0 gateway 10.0.1.1

 

 

I have 4 spare 1gb ports on the server

I need to be able to mount the server on the  10.0.2.x network  without going through the firewall, as the speed through the firewall is horribly slow due to packets being analyzed.

The long term and correct solution seems to be a switch inbetween the firewall and the subnets, however i need a quick fix right now to give a few clients access to the server without hitting the sonicwall.

So i was thinking about running a cat 6 cable from eth1 nic to switch 2 - 10.0.3.x network switch and assigning it an ip 10.0.3.5/255.255.255.0 gateway 10.0.3.1, but then on reading i am wondering wether i need to put the gateway in at all on this connection?

 

However i want to make sure that packets to and from that subnet will only go in and out over that nic and not get lost or try going through the firewall

if i look at the ip route on the server i currently see the following

192.168.100.0/24 dev ib0 proto kernel scope link src 192.168.100.19
10.0.1.0/24 dev eth2 proto kernel scope link src 10.0.1.17
169.254.0.0/16 dev ib0 scope link
default via 10.0.1.1 dev eth2

 

 

i am concerned with trying this and freaking out the network causing broadcast storms etc, so i am looking to clarify what the correct way of doing this would be

 

thanks for any advice

Thread Tags:
  networking 

0 Responses   0 Plus One's   0 Comments  
 

I have two different switches that go to my sonicwall firewall, 

 

switch a is on the 10.0.1.x subnet and switch b is on the 10.0.2.x subnet.

 

On switch A is a Redhat 5.11 File server with static network IP 10.0.1.5 via eth2 which is a 10gbe etehrnet card.

 

right now the server  has 2 active network connections - 

IB0 is the infiniband connection to a separate subnet 192.168.100.x this network is local and doesn't goto the firewall, and has a bunch of servers and the other flames on it used for speedier file transfers for the flames.

 

eth2 which is a 10gbe nic and is currently connected to the switch on 10.0.1.x and gives me access to 

network, 10.0.1.5/255.255.255.0 gateway 10.0.1.1

 

 

I have 4 spare 1gb ports on the server

I need to be able to mount the server on the  10.0.2.x network  without going through the firewall, as the speed through the firewall is horribly slow due to packets being analyzed.

The long term and correct solution seems to be a switch inbetween the firewall and the subnets, however i need a quick fix right now to give a few clients access to the server without hitting the sonicwall.

So i was thinking about running a cat 6 cable from eth1 nic to switch 2 - 10.0.3.x network switch and assigning it an ip 10.0.3.5/255.255.255.0 gateway 10.0.3.1, but then on reading i am wondering wether i need to put the gateway in at all on this connection?

 

However i want to make sure that packets to and from that subnet will only go in and out over that nic and not get lost or try going through the firewall

if i look at the ip route on the server i currently see the following

192.168.100.0/24 dev ib0 proto kernel scope link src 192.168.100.19
10.0.1.0/24 dev eth2 proto kernel scope link src 10.0.1.17
169.254.0.0/16 dev ib0 scope link
default via 10.0.1.1 dev eth2

 

 

i am concerned with trying this and freaking out the network causing broadcast storms etc, so i am looking to clarify what the correct way of doing this would be

 

thanks for any advice

QSFP+ to LC optical breakout?

$
0
0
QSFP+ to LC optical breakout?
posted by Bryce Evans on Jan. 8, 2016, 1:50 p.m. (2 days ago)
Has anyone ever used something like these? (specifically with the newer brocade switches)


I want to break out some 40gig ports to 10gig, without paying insane QSFP+/MPO prices

Brocade actually doesn't make this type of cable yet, so they're gonna give me some hardware to test with.

Reliable vendors/experiences are appreciated!

Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
Has anyone ever used something like these? (specifically with the newer brocade switches)


I want to break out some 40gig ports to 10gig, without paying insane QSFP+/MPO prices

Brocade actually doesn't make this type of cable yet, so they're gonna give me some hardware to test with.

Reliable vendors/experiences are appreciated!

Max number of things in a directory on an Isilon

$
0
0
Max number of things in a directory on an Isilon
posted by Will Rosecrans on Jan. 8, 2016, 6:50 p.m. (2 days ago)
How many is too many? What's the biggest directory you know about on your stack? I am investigating some performance gremlins on an Isilon stack and I am curious what you guys consider best practice in that regard. Using OneFS 7, with NFSv3 mounts.
Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
How many is too many? What's the biggest directory you know about on your stack? I am investigating some performance gremlins on an Isilon stack and I am curious what you guys consider best practice in that regard. Using OneFS 7, with NFSv3 mounts.

Off course this is off topic (David Bowie)

Fortinet ssh backdoor

$
0
0
Fortinet ssh backdoor
posted by Michael Oliver on Jan. 13, 2016, 12:25 a.m.

Looks like fortinet did not want juniper to have all the fun. Good thing is most everyone should be patched up beyond 5.0.7

http://seclists.org/fulldisclosure/2016/Jan/26

Michael Oliver
mcoliver@gmail.com
858.336.1438

Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 

Looks like fortinet did not want juniper to have all the fun. Good thing is most everyone should be patched up beyond 5.0.7

http://seclists.org/fulldisclosure/2016/Jan/26

Michael Oliver
mcoliver@gmail.com
858.336.1438

RV VSync issues

$
0
0
RV VSync issues
posted by Samuel Nair on Jan. 13, 2016, 3:05 a.m. (1 day ago)
I am hoping someone on the list has discovered the magic combination of Card, Driver version, RV Version & settings to solve some of the VSync issues that we are experiencing.

I am testing with:

Centos 6.4 (kernel-2.6.32-431)
Quadro K4000
nvidia-346.35
rv-6.2.6
Dual HP-Z24i 1920x1200 monitors

The problem we are experiencing is,

ensure GL_SYNC_DISPLAY_DEVICE is set to primary monitor
load up a quicktime in RV on primary monitor == no tearing
load up a second instance of RV on the second monitor == tearing on both sessions

The second instance does not even need to have any quicktime loaded, the very act of moving my mouse pointer within the RV display area causes the first instance of RV to tear. Seems like interaction with any openGL app on the second monitor seems to cause RV to tear.

I have followed RV support recommended settings (https://shotgunsoftware.zendesk.com/entries/96048918-V-Sync-Tearing), but no dice.

I am hoping someone on this list has had similar experiences and maybe has the right combination to get this to work

Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
I am hoping someone on the list has discovered the magic combination of Card, Driver version, RV Version & settings to solve some of the VSync issues that we are experiencing.

I am testing with:

Centos 6.4 (kernel-2.6.32-431)
Quadro K4000
nvidia-346.35
rv-6.2.6
Dual HP-Z24i 1920x1200 monitors

The problem we are experiencing is,

ensure GL_SYNC_DISPLAY_DEVICE is set to primary monitor
load up a quicktime in RV on primary monitor == no tearing
load up a second instance of RV on the second monitor == tearing on both sessions

The second instance does not even need to have any quicktime loaded, the very act of moving my mouse pointer within the RV display area causes the first instance of RV to tear. Seems like interaction with any openGL app on the second monitor seems to cause RV to tear.

I have followed RV support recommended settings (https://shotgunsoftware.zendesk.com/entries/96048918-V-Sync-Tearing), but no dice.

I am hoping someone on this list has had similar experiences and maybe has the right combination to get this to work

PCoIP

$
0
0
PCoIP
posted by Peter Daly on Jan. 14, 2016, 8:38 a.m. (1 day ago)

Hey all,

I'm looking into setting up remote access to one of our local Z640 workstations running Centos from another site using a zero client dual screen, mouse,keyboard,wacom.

I've been looking at the pcoip option using the teradici TERA2240 or TERA2220 chipset. I haven't had a chance to see one in action to install or configure yet.

Do I need just a pcie teradici host card plus a zero client like a HP t310?

How is the connection made to the Host Card, does it use VPN would I need to get any specific management software to allow the zero client to access the host card remotely?

 

Thanks

Peter Daly

Windmill Lane VFX

Thread Tags:
  Remote Workstation via Teradici 

0 Responses   3 Plus One's   1 Comments  
 

Hey all,

I'm looking into setting up remote access to one of our local Z640 workstations running Centos from another site using a zero client dual screen, mouse,keyboard,wacom.

I've been looking at the pcoip option using the teradici TERA2240 or TERA2220 chipset. I haven't had a chance to see one in action to install or configure yet.

Do I need just a pcie teradici host card plus a zero client like a HP t310?

How is the connection made to the Host Card, does it use VPN would I need to get any specific management software to allow the zero client to access the host card remotely?

 

Thanks

Peter Daly

Windmill Lane VFX


OpenSSH bug to be announced. Patch now.

$
0
0
OpenSSH bug to be announced. Patch now.
posted by Ken Spickler on Jan. 14, 2016, 11:30 a.m. (1 day ago)
From a post on the OpenBSD Journal (http://undeadly.org/cgi?action=article&sid=20160114142733&mode=expanded), "this is the most serious bug you'll hear about this week".

You can mitigate the risk created by the bug prior to patching as follows:
For OpenSSH >= 5.4 the vulnerable code in the client can be completely disabled by adding 'UseRoaming no' to the global ssh_config(5) file, or to user configuration in ~/.ssh/config, or by passing -oUseRoaming=no on the command line.
Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
From a post on the OpenBSD Journal (http://undeadly.org/cgi?action=article&sid=20160114142733&mode=expanded), "this is the most serious bug you'll hear about this week".

You can mitigate the risk created by the bug prior to patching as follows:
For OpenSSH >= 5.4 the vulnerable code in the client can be completely disabled by adding 'UseRoaming no' to the global ssh_config(5) file, or to user configuration in ~/.ssh/config, or by passing -oUseRoaming=no on the command line.

Mattermost?

$
0
0
Mattermost?
posted by Jean-Francois Panisset on Jan. 14, 2016, 12:35 p.m. (1 day ago)
Has anyone played around with http://www.mattermost.org/ which claims to be a self-hosted Slack-like solution?

People rave about Slack, but it's yet another solution that will cause issues with NDA projects, and it gets pricey very quickly.

JF

Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
Has anyone played around with http://www.mattermost.org/ which claims to be a self-hosted Slack-like solution?

People rave about Slack, but it's yet another solution that will cause issues with NDA projects, and it gets pricey very quickly.

JF

ffmpeg vulnerability

$
0
0
ffmpeg vulnerability
posted by Jean-Francois Panisset on Jan. 15, 2016, 9:40 a.m. (1 day ago)
According to this:

http://news.softpedia.com/news/zero-day-ffmpeg-vulnerability-lets-anyone-steal-files-from-remote-machines-498880.shtml

"ffmpeg has a vulnerability in the current version that allows the attacker to create a specially crafted video file, downloading which will send files from a user PC to a remote attacker server. The attack does not even require the user to open that file - for example, KDE Dolphin thumbnail generation is enough. Desktop search indexers (i.e. baloo) could be affected. ffprobe is affected, basically all operations with file that involve ffmpeg reading it are affected," reads an Arch Linux bug report submitted today."

JF
Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
According to this:

http://news.softpedia.com/news/zero-day-ffmpeg-vulnerability-lets-anyone-steal-files-from-remote-machines-498880.shtml

"ffmpeg has a vulnerability in the current version that allows the attacker to create a specially crafted video file, downloading which will send files from a user PC to a remote attacker server. The attack does not even require the user to open that file - for example, KDE Dolphin thumbnail generation is enough. Desktop search indexers (i.e. baloo) could be affected. ffprobe is affected, basically all operations with file that involve ffmpeg reading it are affected," reads an Arch Linux bug report submitted today."

JF

Optical Flares Scrambled Text Nuke

$
0
0
Optical Flares Scrambled Text Nuke
posted by RobertC on Jan. 15, 2016, 8:40 p.m. (1 day ago)

Hey,

We've been running into issues recently where when you load a nuke script that has an Optical Flares (video copilot) node inside, the node tree will show scrambled text. This only affects a handful of random machines. If the user opened a blank Nuke script, and copied and pasted the affected script into the blank script, the text would no longer be scrambled. Our pipeline department has found that the scrambled text issue always occurs if the Optical Flares plugin is loaded before the GUI. All of our machines are on the simular hardware, so no differences between a machine that has the issue and one that doesn't.

We've reached out to Video Copilot to see if they are able to reproduce the issue but we can't get a response from their support.

Is anyone else having this issue?

 

Example Screenshot of the issue.

Thread Tags:
  plugins 

0 Responses   0 Plus One's   0 Comments  
 

Hey,

We've been running into issues recently where when you load a nuke script that has an Optical Flares (video copilot) node inside, the node tree will show scrambled text. This only affects a handful of random machines. If the user opened a blank Nuke script, and copied and pasted the affected script into the blank script, the text would no longer be scrambled. Our pipeline department has found that the scrambled text issue always occurs if the Optical Flares plugin is loaded before the GUI. All of our machines are on the simular hardware, so no differences between a machine that has the issue and one that doesn't.

We've reached out to Video Copilot to see if they are able to reproduce the issue but we can't get a response from their support.

Is anyone else having this issue?

 

Example Screenshot of the issue.

chef/puppet/ansible/salt and autodesk maya and windows 10

$
0
0
chef/puppet/ansible/salt and autodesk maya and windows 10
posted by Steve Molin on Jan. 18, 2016, 8:06 p.m. (1 day ago)

Okay, flame me for a newbie question if you have to, but any thoughts on what's the easiest way to deploy Maya onto Windows 10? My web-search-fu has failed me. No preference for agent/agentless, and I do have Linux machines handy for a master/controller.

Thanks!


0 Responses   0 Plus One's   0 Comments  
 

Okay, flame me for a newbie question if you have to, but any thoughts on what's the easiest way to deploy Maya onto Windows 10? My web-search-fu has failed me. No preference for agent/agentless, and I do have Linux machines handy for a master/controller.

Thanks!

Videocopilot - Optical Flares 1.0.8 on a renderfarm....

$
0
0
Videocopilot - Optical Flares 1.0.8 on a renderfarm....
posted by Dan Young on Jan. 19, 2016, 1 p.m. (1 day ago)
Hello everyone,

Seeing weird activity from their product - and they don't seem to know how to fix it. We are getting frames rendered through, but it exits with a nice 139 code, which pisses everyone and everything involved post rendered frame off.

Their support team has just stopped responding, because this is probably well beyond out of their technical depth - has anyone seen this? We started with errors because we didn't have X11 sessions on our renderfarm, because, well, nobody runs X on their farm - but even with GPU resources turned off, it's still attempting to use X and will 139 without it.

Anyone come up against it? Yes, the GPU disable env var is set, and the logs maintain that it's not attempting a GPU render, so it's not that. Sample renderlog;

---

--[ Begin f#753132:0 on Wed Jan 13 14:46:42 2016

--[ Host is rb013

--[ Using /bin/tcsh to execute go /job/comms/heineken_barlights_cityreflect_3000873// && cd /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile && nuke-progress env NUKE_PATH=${NUKE_PATH}:/job/fcomms/pipeline/2015.1/tk-core/install/apps/app_store/tk-nuke-writenode/v1.0.18/gizmos CM_NO_SGTK=1 nuke -c `nmem -m -f 0.75`M -m `nproc` -t -X "ShotgunWrite1" -f /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile/snapshot/mpascuzz_comp_test_v001.2016-01-13-14-46-35.nk 0001 0002 0003 0004 0005 0006 0007 0008 0009 0010

Nuke 9.0v7, 64 bit, built Aug 17 2015.

Copyright (c) 2015 The Foundry Visionmongers Ltd. All Rights Reserved.

pgBokeh v1.4.2, built on Nov 18 2014 at 20:55:55

(c) 2011 - Present by Peregrine Labs, a division of Peregrine Visual Storytelling Ltd. All rights reserved.

INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap

INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap

INFO: [commsnukelib.startup] init_all took 0.55 seconds

*OF* Logging Enabled

*OF* Welcome to OpticalFlares for NUKE 1.0.8, Instance ID: 0

*OF* Creating Instance: 0 **********

*OF* GPU rendering disabled

*OF* Unable to init X11!*OF* OF/nuke is expecting Presets and Textures in: /net/global/opt/videocopilot/opticalflares-1.0.8/nuke-8.0v6

*OpticalFlares* Unable to get X11 support!

*OF* Renderer has error: 0*OF* OF/nuke is expecting License Server at: 172.17.1.6:47200

*OF* License Acquired from 172.17.1.6:47200

*OF* NODE: 0 + attach()

......

Writing.......... (frames are written)
..............

*OF* Destroying instance: 0

/job/fcomms/pipeline/2015.1/common/bin/nuke-progress: line 38: 95707 Segmentation fault (core dumped) stdbuf -o0 "$@" > "$OUT"

--[ Command returned exit code 139

--[ Completed f#753132:0 on Wed Jan 13 14:46:48 2016

--[ Host was rb013
---

Any thoughts appreciated, this is a very silly problem to have. Surely someone is rendering OF in Nuke under Linux on a non-X11 renderfarm....

DY

--

Framestore
Dan Young Lead Systems Engineer
LondonNew YorkLos AngelesMontral
T+1 212 775 0600
135 Spring Street, New York NY 10012
TwitterFacebookframestore.com
Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
Hello everyone,

Seeing weird activity from their product - and they don't seem to know how to fix it. We are getting frames rendered through, but it exits with a nice 139 code, which pisses everyone and everything involved post rendered frame off.

Their support team has just stopped responding, because this is probably well beyond out of their technical depth - has anyone seen this? We started with errors because we didn't have X11 sessions on our renderfarm, because, well, nobody runs X on their farm - but even with GPU resources turned off, it's still attempting to use X and will 139 without it.

Anyone come up against it? Yes, the GPU disable env var is set, and the logs maintain that it's not attempting a GPU render, so it's not that. Sample renderlog;

---

--[ Begin f#753132:0 on Wed Jan 13 14:46:42 2016

--[ Host is rb013

--[ Using /bin/tcsh to execute go /job/comms/heineken_barlights_cityreflect_3000873// && cd /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile && nuke-progress env NUKE_PATH=${NUKE_PATH}:/job/fcomms/pipeline/2015.1/tk-core/install/apps/app_store/tk-nuke-writenode/v1.0.18/gizmos CM_NO_SGTK=1 nuke -c `nmem -m -f 0.75`M -m `nproc` -t -X "ShotgunWrite1" -f /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile/snapshot/mpascuzz_comp_test_v001.2016-01-13-14-46-35.nk 0001 0002 0003 0004 0005 0006 0007 0008 0009 0010

Nuke 9.0v7, 64 bit, built Aug 17 2015.

Copyright (c) 2015 The Foundry Visionmongers Ltd. All Rights Reserved.

pgBokeh v1.4.2, built on Nov 18 2014 at 20:55:55

(c) 2011 - Present by Peregrine Labs, a division of Peregrine Visual Storytelling Ltd. All rights reserved.

INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap

INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap

INFO: [commsnukelib.startup] init_all took 0.55 seconds

*OF* Logging Enabled

*OF* Welcome to OpticalFlares for NUKE 1.0.8, Instance ID: 0

*OF* Creating Instance: 0 **********

*OF* GPU rendering disabled

*OF* Unable to init X11!*OF* OF/nuke is expecting Presets and Textures in: /net/global/opt/videocopilot/opticalflares-1.0.8/nuke-8.0v6

*OpticalFlares* Unable to get X11 support!

*OF* Renderer has error: 0*OF* OF/nuke is expecting License Server at: 172.17.1.6:47200

*OF* License Acquired from 172.17.1.6:47200

*OF* NODE: 0 + attach()

......

Writing.......... (frames are written)
..............

*OF* Destroying instance: 0

/job/fcomms/pipeline/2015.1/common/bin/nuke-progress: line 38: 95707 Segmentation fault (core dumped) stdbuf -o0 "$@" > "$OUT"

--[ Command returned exit code 139

--[ Completed f#753132:0 on Wed Jan 13 14:46:48 2016

--[ Host was rb013
---

Any thoughts appreciated, this is a very silly problem to have. Surely someone is rendering OF in Nuke under Linux on a non-X11 renderfarm....

DY

--

Framestore
Dan Young Lead Systems Engineer
LondonNew YorkLos AngelesMontral
T+1 212 775 0600
135 Spring Street, New York NY 10012
TwitterFacebookframestore.com

Videocopilot - Optical Flares 1.0.8 on a renderfarm....

$
0
0
Videocopilot - Optical Flares 1.0.8 on a renderfarm....
posted by Rory Falloon on Jan. 19, 2016, 1:30 p.m. (1 day ago)
That's very interesting - I'd never considered that.

Our blades do not *run* X but they are a direct image clone of our workstation image (so as to keep all libs the same), so they have all relevant X libs installed.

We have to set os.environ["OPTICAL_FLARES_NO_GPU"] = "true" to get a successful render, but I figured that's all it needed.


Rory Falloon
Snr. Systems Administrator
Mr. X Inc.
T:  (416) 595-6222, x 290


From: "Chris Hyman" <chris@cyberdyne-services.com>
To: studiosysadmins-discuss@studiosysadmins.com
Sent: Tuesday, January 19, 2016 1:19:17 PM
Subject: Re: [SSA-Discuss] Videocopilot - Optical Flares 1.0.8 on a        renderfarm....

Hi Dan,

Yes, you need to fake X  in some way without it it will not render.

We use xvfb to fake an X11 server

You also may require to install NVidia drivers (for some libraries video copilot requires).

Enjoy installing xvfb, or installing X :-)


Chris
On 19 Jan 2016, at 17:55, Dan Young <dan.young@framestore.com> wrote:

Hello everyone,

Seeing weird activity from their product - and they don't seem to know how to fix it. We are getting frames rendered through, but it exits with a nice 139 code, which pisses everyone and everything involved post rendered frame off.

Their support team has just stopped responding, because this is probably well beyond out of their technical depth - has anyone seen this? We started with errors because we didn't have X11 sessions on our renderfarm, because, well, nobody runs X on their farm - but even with GPU resources turned off, it's still attempting to use X and will 139 without it.

Anyone come up against it? Yes, the GPU disable env var is set, and the logs maintain that it's not attempting a GPU render, so it's not that.  Sample renderlog;

---
--[ Begin f#753132:0 on Wed Jan 13 14:46:42 2016
--[ Host is rb013
--[ Using /bin/tcsh to execute go /job/comms/heineken_barlights_cityreflect_3000873// && cd /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile && nuke-progress env NUKE_PATH=${NUKE_PATH}:/job/fcomms/pipeline/2015.1/tk-core/install/apps/app_store/tk-nuke-writenode/v1.0.18/gizmos CM_NO_SGTK=1 nuke -c `nmem -m -f 0.75`M -m `nproc` -t -X "ShotgunWrite1" -f /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile/snapshot/mpascuzz_comp_test_v001.2016-01-13-14-46-35.nk 0001 0002 0003 0004 0005 0006 0007 0008 0009 0010
Nuke 9.0v7, 64 bit, built Aug 17 2015.
Copyright (c) 2015 The Foundry Visionmongers Ltd. All Rights Reserved.
pgBokeh v1.4.2, built on Nov 18 2014 at 20:55:55
(c) 2011 - Present by Peregrine Labs, a division of Peregrine Visual Storytelling Ltd. All rights reserved.
INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap
INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap
INFO: [commsnukelib.startup] init_all took 0.55 seconds
*OF* Logging Enabled
*OF* Welcome to OpticalFlares for NUKE 1.0.8, Instance ID: 0
*OF* Creating Instance: 0 **********
*OF* GPU rendering disabled
*OF* Unable to init X11!*OF* OF/nuke is expecting Presets and Textures in: /net/global/opt/videocopilot/opticalflares-1.0.8/nuke-8.0v6
*OpticalFlares* Unable to get X11 support!
*OF* Renderer has error: 0*OF* OF/nuke is expecting License Server at: 172.17.1.6:47200
*OF* License Acquired from 172.17.1.6:47200
*OF* NODE: 0 + attach()
......
Writing.......... (frames are written)
..............
*OF* Destroying instance: 0
/job/fcomms/pipeline/2015.1/common/bin/nuke-progress: line 38: 95707 Segmentation fault (core dumped) stdbuf -o0 "$@" > "$OUT"
--[ Command returned exit code 139
--[ Completed f#753132:0 on Wed Jan 13 14:46:48 2016
--[ Host was rb013
---

Any thoughts appreciated, this is a very silly problem to have. Surely someone is rendering OF in Nuke under Linux on a non-X11 renderfarm....

DY

--

Framestore
 
Dan Young Lead Systems Engineer
 
LondonNew YorkLos AngelesMontral
T  +1 212 775 0600
135 Spring Street, New York NY 10012
TwitterFacebookframestore.com
To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe


To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe

Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
That's very interesting - I'd never considered that.

Our blades do not *run* X but they are a direct image clone of our workstation image (so as to keep all libs the same), so they have all relevant X libs installed.

We have to set os.environ["OPTICAL_FLARES_NO_GPU"] = "true" to get a successful render, but I figured that's all it needed.


Rory Falloon
Snr. Systems Administrator
Mr. X Inc.
T:  (416) 595-6222, x 290


From: "Chris Hyman" <chris@cyberdyne-services.com>
To: studiosysadmins-discuss@studiosysadmins.com
Sent: Tuesday, January 19, 2016 1:19:17 PM
Subject: Re: [SSA-Discuss] Videocopilot - Optical Flares 1.0.8 on a        renderfarm....

Hi Dan,

Yes, you need to fake X  in some way without it it will not render.

We use xvfb to fake an X11 server

You also may require to install NVidia drivers (for some libraries video copilot requires).

Enjoy installing xvfb, or installing X :-)


Chris
On 19 Jan 2016, at 17:55, Dan Young <dan.young@framestore.com> wrote:

Hello everyone,

Seeing weird activity from their product - and they don't seem to know how to fix it. We are getting frames rendered through, but it exits with a nice 139 code, which pisses everyone and everything involved post rendered frame off.

Their support team has just stopped responding, because this is probably well beyond out of their technical depth - has anyone seen this? We started with errors because we didn't have X11 sessions on our renderfarm, because, well, nobody runs X on their farm - but even with GPU resources turned off, it's still attempting to use X and will 139 without it.

Anyone come up against it? Yes, the GPU disable env var is set, and the logs maintain that it's not attempting a GPU render, so it's not that.  Sample renderlog;

---
--[ Begin f#753132:0 on Wed Jan 13 14:46:42 2016
--[ Host is rb013
--[ Using /bin/tcsh to execute go /job/comms/heineken_barlights_cityreflect_3000873// && cd /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile && nuke-progress env NUKE_PATH=${NUKE_PATH}:/job/fcomms/pipeline/2015.1/tk-core/install/apps/app_store/tk-nuke-writenode/v1.0.18/gizmos CM_NO_SGTK=1 nuke -c `nmem -m -f 0.75`M -m `nproc` -t -X "ShotgunWrite1" -f /job/comms/heineken_barlights_cityreflect_3000873/sandbox/mpascuzz/comp/comp/work/nuke/workfile/snapshot/mpascuzz_comp_test_v001.2016-01-13-14-46-35.nk 0001 0002 0003 0004 0005 0006 0007 0008 0009 0010
Nuke 9.0v7, 64 bit, built Aug 17 2015.
Copyright (c) 2015 The Foundry Visionmongers Ltd. All Rights Reserved.
pgBokeh v1.4.2, built on Nov 18 2014 at 20:55:55
(c) 2011 - Present by Peregrine Labs, a division of Peregrine Visual Storytelling Ltd. All rights reserved.
INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap
INFO: [commslib.shotgun.engine] Skipping SGTK bootstrap
INFO: [commsnukelib.startup] init_all took 0.55 seconds
*OF* Logging Enabled
*OF* Welcome to OpticalFlares for NUKE 1.0.8, Instance ID: 0
*OF* Creating Instance: 0 **********
*OF* GPU rendering disabled
*OF* Unable to init X11!*OF* OF/nuke is expecting Presets and Textures in: /net/global/opt/videocopilot/opticalflares-1.0.8/nuke-8.0v6
*OpticalFlares* Unable to get X11 support!
*OF* Renderer has error: 0*OF* OF/nuke is expecting License Server at: 172.17.1.6:47200
*OF* License Acquired from 172.17.1.6:47200
*OF* NODE: 0 + attach()
......
Writing.......... (frames are written)
..............
*OF* Destroying instance: 0
/job/fcomms/pipeline/2015.1/common/bin/nuke-progress: line 38: 95707 Segmentation fault (core dumped) stdbuf -o0 "$@" > "$OUT"
--[ Command returned exit code 139
--[ Completed f#753132:0 on Wed Jan 13 14:46:48 2016
--[ Host was rb013
---

Any thoughts appreciated, this is a very silly problem to have. Surely someone is rendering OF in Nuke under Linux on a non-X11 renderfarm....

DY

--

Framestore
 
Dan Young Lead Systems Engineer
 
LondonNew YorkLos AngelesMontral
T  +1 212 775 0600
135 Spring Street, New York NY 10012
TwitterFacebookframestore.com
To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe


To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe


chef/puppet/ansible/salt and autodesk maya and windows 10

$
0
0
chef/puppet/ansible/salt and autodesk maya and windows 10
posted by Michael Oliver on Jan. 19, 2016, 3:50 p.m. (1 day ago)
how many machines? Do you have a deployment solution already? Are you running local installs or putting it on a network share?

You can always create a deployment using the maya installer and then run a powershell "for" loop (or create a gpo) to push it out to all your workstations. That is probably the "easiest" if you are starting from nothing and want local installs. After you are going to want some sort of launcher to run a single usersetup.py file so that you can standardize plugins, shelves, etc.. on a shared drive.

Rough example of a looped local install:

That being said you are going to want to setup some sort of deployment/standardization solution as mentioned in your subject but that is a bigger can of worms with each having their own set of pros and cons.

Search the sysadmins list for a few threads where they are debated/compared.

On Mon, Jan 18, 2016 at 5:06 PM, Steve Molin <content@studiosysadmins.com> wrote:

Okay, flame me for a newbie question if you have to, but any thoughts on what's the easiest way to deploy Maya onto Windows 10? My web-search-fu has failed me. No preference for agent/agentless, and I do have Linux machines handy for a master/controller.

Thanks!


To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe



--
Michael Oliver
mcoliver@gmail.com
858.336.1438
PGP Public Key: https://goo.gl/a2CnHF
Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
how many machines? Do you have a deployment solution already? Are you running local installs or putting it on a network share?

You can always create a deployment using the maya installer and then run a powershell "for" loop (or create a gpo) to push it out to all your workstations. That is probably the "easiest" if you are starting from nothing and want local installs. After you are going to want some sort of launcher to run a single usersetup.py file so that you can standardize plugins, shelves, etc.. on a shared drive.

Rough example of a looped local install:

That being said you are going to want to setup some sort of deployment/standardization solution as mentioned in your subject but that is a bigger can of worms with each having their own set of pros and cons.

Search the sysadmins list for a few threads where they are debated/compared.

On Mon, Jan 18, 2016 at 5:06 PM, Steve Molin <content@studiosysadmins.com> wrote:

Okay, flame me for a newbie question if you have to, but any thoughts on what's the easiest way to deploy Maya onto Windows 10? My web-search-fu has failed me. No preference for agent/agentless, and I do have Linux machines handy for a master/controller.

Thanks!


To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe



--
Michael Oliver
mcoliver@gmail.com
858.336.1438
PGP Public Key: https://goo.gl/a2CnHF

Atomic Fiction Oakland is Hiring

$
0
0
Atomic Fiction Oakland is Hiring
posted by Shawn Wallbridge on Jan. 20, 2016, 12:50 p.m. (1 day ago)
Atomic Fiction is looking for a talented Jr. Systems Administrator!

Applicants should have a passion for systems engineering and systems architecture, and enjoy figuring out how and why things work. Server uptime and artist efficiency / happiness are defining goals in this role. Day-to-day responsibilities include support and maintenance of our workstations, network infrastructure and file servers. Of course, youll also be instrumental in planning the future expansion of the studio.

Ideal Candidates will possess the following:
  • 1-3 years in a systems administration role
  • Enjoys working with and helping others
  • Comfortable in Osx, Linux and Windows
  • Strong problem solving skills
  • Passionate about working in a startup environment at the intersection of art and technology

Nice to have:

  • Experience with modern configuration management systems (Puppet, Chef, Ansible)
  • Working knowledge of automation languages such as BASH and Python
  • Comfortable with cloud services such as Amazon Web Services and Google Compute Engine.
If you are interested, or can recommend someone, please email me.

thanks
shawn


Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
Atomic Fiction is looking for a talented Jr. Systems Administrator!

Applicants should have a passion for systems engineering and systems architecture, and enjoy figuring out how and why things work. Server uptime and artist efficiency / happiness are defining goals in this role. Day-to-day responsibilities include support and maintenance of our workstations, network infrastructure and file servers. Of course, youll also be instrumental in planning the future expansion of the studio.

Ideal Candidates will possess the following:
  • 1-3 years in a systems administration role
  • Enjoys working with and helping others
  • Comfortable in Osx, Linux and Windows
  • Strong problem solving skills
  • Passionate about working in a startup environment at the intersection of art and technology

Nice to have:

  • Experience with modern configuration management systems (Puppet, Chef, Ansible)
  • Working knowledge of automation languages such as BASH and Python
  • Comfortable with cloud services such as Amazon Web Services and Google Compute Engine.
If you are interested, or can recommend someone, please email me.

thanks
shawn


Mid week fun

ingest & edit for a live event

$
0
0
ingest & edit for a live event
posted by Rob LaRose on Jan. 20, 2016, 8:10 p.m. (1 day ago)

Hi folks,

Anybody ever done ingest & edit of a live event, or know a company that rolls this solution into a room in a rack?

Need to ingest a single live video stream, have 2-3 editors and graphics people be able to edit together bits as its happening for export to social media platforms.

I dont want to invent this myself, since the event is live.. Id like very much to talk to someone who builds/provides this setup for a living and I just have to bring my editors.

Rob






rob larose | engineer | rock paper scissors | 212-255-6446 | www.rockpaperscissors.com

Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 

Hi folks,

Anybody ever done ingest & edit of a live event, or know a company that rolls this solution into a room in a rack?

Need to ingest a single live video stream, have 2-3 editors and graphics people be able to edit together bits as its happening for export to social media platforms.

I dont want to invent this myself, since the event is live.. Id like very much to talk to someone who builds/provides this setup for a living and I just have to bring my editors.

Rob






rob larose | engineer | rock paper scissors | 212-255-6446 | www.rockpaperscissors.com

Lightweight virus endpoint software for artist/edit stations

$
0
0
Lightweight virus endpoint software for artist/edit stations
posted by Julian Firminger on Jan. 21, 2016, 11 a.m. (1 day ago)
Gents,

We're reviewing what vendor to use for endpoint protection for artist/edit systems that operate outside our doors or on remote, non-internet secure sites. What are people using? (windows clearly)

Specifically I'm interested in knowing if anyone has compared the performance drag of different vendors in relation to running alongside an NLE or VFX software.

Julian Firminger

Snr. Systems Administrator,
United Broadcast Facilities
Amsterdam, The Netherlands
Thread Tags:
  discuss-at-studiosysadmins 

0 Responses   0 Plus One's   0 Comments  
 
Gents,

We're reviewing what vendor to use for endpoint protection for artist/edit systems that operate outside our doors or on remote, non-internet secure sites. What are people using? (windows clearly)

Specifically I'm interested in knowing if anyone has compared the performance drag of different vendors in relation to running alongside an NLE or VFX software.

Julian Firminger

Snr. Systems Administrator,
United Broadcast Facilities
Amsterdam, The Netherlands
Viewing all 3749 articles
Browse latest View live




Latest Images