BGP routers / vyos.

February 19, 2016, 11:50 am

≫ Next: Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

≪ Previous: Apple Device Enrolment Program

$

0

0

BGP routers / vyos.

posted by Greg Whynott on Feb. 19, 2016, 2:50 p.m. (1 day ago)

I'll soon be deploying BGP between us and our providers, the requirements are 1 10 gig interface facing inside and 3 1 gig outbound facing interfaces. Last time I did this I was told budget was a non issue (working for the government) and bought a few ASR's from Cisco. This time round budget is a factor and when I mentioned "might be 50k or so" jaws dropped. So now i'm considering alternative options. It needs to be fully supported, not a DYI hopefully but that isn't precluded as of yet. Thus far I have looked at Ubiquiti networks (ERPro-8) but it doesn't have a 10 gig link. I'd like to avoid link aggregation if possible as it doesn't work ideally most of the time in these situations. I'm now looking at vyOS.. Anyone using that or similar instead of 'big iron' solutions for their pure routing needs? -greg

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

I'll soon be deploying BGP between us and our providers, the requirements are 1 10 gig interface facing inside and 3 1 gig outbound facing interfaces. Last time I did this I was told budget was a non issue (working for the government) and bought a few ASR's from Cisco. This time round budget is a factor and when I mentioned "might be 50k or so" jaws dropped. So now i'm considering alternative options. It needs to be fully supported, not a DYI hopefully but that isn't precluded as of yet. Thus far I have looked at Ubiquiti networks (ERPro-8) but it doesn't have a 10 gig link. I'd like to avoid link aggregation if possible as it doesn't work ideally most of the time in these situations. I'm now looking at vyOS.. Anyone using that or similar instead of 'big iron' solutions for their pure routing needs? -greg

---

Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

February 21, 2016, 8:45 am

≫ Next: Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

≪ Previous: BGP routers / vyos.

$

0

0

Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

posted by Dave Lajoie on Feb. 21, 2016, 11:45 a.m.

Hello Guys, I have been using Nexenta for almost 3 years, but we ran into a serious issue, where Nexenta support simply cannot resolve. Basically for months, we have random files disappear. From NFS perspective, the file is present, however the permissions are all ???-???-???. If we attempt to delete that file, we get a message the file cannot be delete. If we go directly on the nexenta server, the file is not even in the zpool. Attempting to delete the file while being logged in on the server, will cause the server to kernel panic. Normally no user can reproduce this kernel panic, somehow being sheilded by the NFS, however if a user attempt to recursively remove a directory ( thru NFS ) - which happen to contain one of those bogus file - the server will crash with a kernel panic. Q1) Has someone experienced similar problem, and how did you resolve it? Q2) It is possible the zpool is corrupted? if so, who to determine if zpool is really healthy? Any help would be appreciated. Best Dave.

Thread Tags:   Nexenta Server Crash

0 Responses     0 Plus One's     0 Comments

 

Hello Guys, I have been using Nexenta for almost 3 years, but we ran into a serious issue, where Nexenta support simply cannot resolve. Basically for months, we have random files disappear. From NFS perspective, the file is present, however the permissions are all ???-???-???. If we attempt to delete that file, we get a message the file cannot be delete. If we go directly on the nexenta server, the file is not even in the zpool. Attempting to delete the file while being logged in on the server, will cause the server to kernel panic. Normally no user can reproduce this kernel panic, somehow being sheilded by the NFS, however if a user attempt to recursively remove a directory ( thru NFS ) - which happen to contain one of those bogus file - the server will crash with a kernel panic. Q1) Has someone experienced similar problem, and how did you resolve it? Q2) It is possible the zpool is corrupted? if so, who to determine if zpool is really healthy? Any help would be appreciated. Best Dave.

Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

February 22, 2016, 5:55 am

≫ Next: Oracle ZFS Customer Satisfaction

≪ Previous: Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

$

0

0

Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

posted by Bobby Jain on Feb. 22, 2016, 8:55 a.m. (2 days ago)

Having dealt with Nexenta tech support a lot, my first piece of advice, is if you have not already done so, get a VM, and recreate the structure/environment the best you can. If the problem is this consistent on the production server, then you should be able to recreate it on the test server. Secondly, which specific version are you on? I was experiencing Nexenta kernal panics that support was never able to help resolve, but was magically fixed in a later release (nothing in the release notes of course). You could try installing a newer version using a snapshot and test if the bug still exists. You can easily restart between versions using the snapshots. What happens when you copy a directory with a problem file locally? Are the files present or missing? On Sun, Feb 21, 2016 at 11:45 AM, Dave Lajoie <content@studiosysadmins.com> wrote: Hello Guys, I have been using Nexenta for almost 3 years, but we ran into a serious issue, where Nexenta support simply cannot resolve. Basically for months, we have random files disappear. From NFS perspective, the file is present, however the permissions are all ???-???-???.If we attempt to delete that file, we get a message the file cannot be delete.If we go directly on the nexenta server, the file is not even in the zpool. Attempting to delete the file while being logged in on the server, will cause the server to kernel panic. Normally no user can reproduce this kernel panic, somehow being sheilded by the NFS, however if a user attempt to recursively remove a directory ( thru NFS ) - which happen to contain one of those bogus file - the server will crash with a kernel panic. Q1) Has someone experienced similar problem, and how did you resolve it? Q2) It is possible the zpool is corrupted? if so, who to determine if zpool is really healthy? Any help would be appreciated. Best Dave. 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     1 Comments

 

Having dealt with Nexenta tech support a lot, my first piece of advice, is if you have not already done so, get a VM, and recreate the structure/environment the best you can. If the problem is this consistent on the production server, then you should be able to recreate it on the test server. Secondly, which specific version are you on? I was experiencing Nexenta kernal panics that support was never able to help resolve, but was magically fixed in a later release (nothing in the release notes of course). You could try installing a newer version using a snapshot and test if the bug still exists. You can easily restart between versions using the snapshots. What happens when you copy a directory with a problem file locally? Are the files present or missing? On Sun, Feb 21, 2016 at 11:45 AM, Dave Lajoie <content@studiosysadmins.com> wrote: Hello Guys, I have been using Nexenta for almost 3 years, but we ran into a serious issue, where Nexenta support simply cannot resolve. Basically for months, we have random files disappear. From NFS perspective, the file is present, however the permissions are all ???-???-???.If we attempt to delete that file, we get a message the file cannot be delete.If we go directly on the nexenta server, the file is not even in the zpool. Attempting to delete the file while being logged in on the server, will cause the server to kernel panic. Normally no user can reproduce this kernel panic, somehow being sheilded by the NFS, however if a user attempt to recursively remove a directory ( thru NFS ) - which happen to contain one of those bogus file - the server will crash with a kernel panic. Q1) Has someone experienced similar problem, and how did you resolve it? Q2) It is possible the zpool is corrupted? if so, who to determine if zpool is really healthy? Any help would be appreciated. Best Dave. To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe

Oracle ZFS Customer Satisfaction

February 22, 2016, 6:07 pm

≫ Next: Qumulo - Max Q Vancouver 3/3/16

≪ Previous: Kernel Panic Crash with Nexenta 4.x / Files which cannot be deleted

$

0

0

Oracle ZFS Customer Satisfaction

posted by Jeff Yana on Feb. 22, 2016, 9:07 p.m. (2 days ago)

Hello All- Given the small installed user base, this question might be somewhat premature, but I am looking for some community feedback regarding Oracle ZFS by way of general impressions and even recommendations, where applicable. I am curious to know if there any shops out there currently using this product as their primary storage solution and whether or not they are generally happy with it? This is not a marketing or sales study but just trying to establish a use case for this product. More to the point, I am curious about the following: How are you using it: As a general filer? As a production server supporting many 10s/100s of artists and compute nodes. - How many clients exactly? - Are you using any NAS acceleration in front of it? As a frame store for high resolution, uncompressed media playback? Or any combination of the above. If as a production server, how well does it handle mixed workloads, ranging from 2D to 3D rendering to large block, sequential IO? General impressions: Where do you think this product shines and where does if fall flat; that is, what does it do well and what does it do not so well?  If purchased recently, did your Oracle ZFS appliance replace a more expensive, traditional NAS or scale-out NAS solution, and if so, which product was that? Would you say the migration was successful, exceeding or meeting most, if not all, of your expectations, or was it unsuccessful and are you sorry that you ever made the move? Support: Oracle’s costly support plans is the butt of all kinds of industry jokes. That said, are you satisfied with the quality of support, do you think you are getting your money’s worth, etc. I have no previous experience using Oracle’s gear outside of what used to be known as SUN servers and storage. I know that many SUN customers saw their support costs double folowing the acquisition (myself included). That was a huge deterrent for me for ever going back to Oracle after that. Thanks in advance for your (brutally honest, but fair) comments and feedback. Jeff

Thread Tags:   ZFS. Storage

0 Responses     0 Plus One's     1 Comments

 

Hello All- Given the small installed user base, this question might be somewhat premature, but I am looking for some community feedback regarding Oracle ZFS by way of general impressions and even recommendations, where applicable. I am curious to know if there any shops out there currently using this product as their primary storage solution and whether or not they are generally happy with it? This is not a marketing or sales study but just trying to establish a use case for this product. More to the point, I am curious about the following: How are you using it: As a general filer? As a production server supporting many 10s/100s of artists and compute nodes. - How many clients exactly? - Are you using any NAS acceleration in front of it? As a frame store for high resolution, uncompressed media playback? Or any combination of the above. If as a production server, how well does it handle mixed workloads, ranging from 2D to 3D rendering to large block, sequential IO? General impressions: Where do you think this product shines and where does if fall flat; that is, what does it do well and what does it do not so well?  If purchased recently, did your Oracle ZFS appliance replace a more expensive, traditional NAS or scale-out NAS solution, and if so, which product was that? Would you say the migration was successful, exceeding or meeting most, if not all, of your expectations, or was it unsuccessful and are you sorry that you ever made the move? Support: Oracle’s costly support plans is the butt of all kinds of industry jokes. That said, are you satisfied with the quality of support, do you think you are getting your money’s worth, etc. I have no previous experience using Oracle’s gear outside of what used to be known as SUN servers and storage. I know that many SUN customers saw their support costs double folowing the acquisition (myself included). That was a huge deterrent for me for ever going back to Oracle after that. Thanks in advance for your (brutally honest, but fair) comments and feedback. Jeff

Qumulo - Max Q Vancouver 3/3/16

February 23, 2016, 3:25 pm

≫ Next: Reminder, this is Tomorrow: Montreal SSA Event: CEV/Qumulo Beer & Banter

≪ Previous: Oracle ZFS Customer Satisfaction

$

0

0

Qumulo - Max Q Vancouver 3/3/16

posted by Anh Quach on Feb. 23, 2016, 6:25 p.m. (1 day ago)

Vancouver community members, Our next Max Q User Group meeting is at CRAFT brewing in the Olympic Village on March 3, 2016, 6PM. Some of your are already registered, but we'd love to welcome anyone else in the community who would like to attend. Follow this link to register: https://maxqvancouver.splashthat.com/ A reminder that Max Q is aQumulotechnical user group, and as such will not be a sales or marketing event. Translate: no sales reps. Members of Qumulo's Product Management team as well as technical staff will be present for discussion. If anyone in the community has anything they'd like to present on, give me a shout directly. Thanks and hope to see you there! -Anh Anh Quach Director, Customer Success ___________________________________ Qumulo, Inc. P:206.260.3588//M:206.949.8841 World's First Data-Aware Scale-Out NAS Twitter///LinkedIn///Facebook///YouTube

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

Vancouver community members, Our next Max Q User Group meeting is at CRAFT brewing in the Olympic Village on March 3, 2016, 6PM. Some of your are already registered, but we'd love to welcome anyone else in the community who would like to attend. Follow this link to register: https://maxqvancouver.splashthat.com/ A reminder that Max Q is aQumulotechnical user group, and as such will not be a sales or marketing event. Translate: no sales reps. Members of Qumulo's Product Management team as well as technical staff will be present for discussion. If anyone in the community has anything they'd like to present on, give me a shout directly. Thanks and hope to see you there! -Anh Anh Quach Director, Customer Success ___________________________________ Qumulo, Inc. P:206.260.3588//M:206.949.8841 World's First Data-Aware Scale-Out NAS Twitter///LinkedIn///Facebook///YouTube

Reminder, this is Tomorrow: Montreal SSA Event: CEV/Qumulo Beer & Banter

February 24, 2016, 7:10 am

≫ Next: Flame 2016 + Z840 + Background reactor - PCoIP extenders?

≪ Previous: Qumulo - Max Q Vancouver 3/3/16

$

0

0

Reminder, this is Tomorrow: Montreal SSA Event: CEV/Qumulo Beer & Banter

posted by  on Feb. 24, 2016, 10:10 a.m.

Hi everyone in the Montreal area, This is a reminder that this event is tomorrow! Youre invited to join us on February 25th, 2016, for a Beer & Banter presentation of the Qumulo Storage Solutions. This is a radical redesign of what storage solutions offer. With Qumulo, you can instantly understand: what files or applications are impacting capacity or performance. what data is hot, warm or cold. how 2015 data growth and performance trends inform your 2016 plans. RSVP Seating is limited. RSVP without delay to insure you see the latest solutions and talk directly to our product specialists. To RSVP, click here Where Le Ballpark - 6660 Rue Clark, Montral, QC H2S, Canada When Thursday, February 25th at 5:30pm We look forward to seeing you there! Link to Event:http://www.studiosysadmins.com/events/view/131/ -- Francis Provencher StudioSysAdmins Montreal Meetings Coordinator <francisp@gmail.com>

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

Hi everyone in the Montreal area, This is a reminder that this event is tomorrow! Youre invited to join us on February 25th, 2016, for a Beer & Banter presentation of the Qumulo Storage Solutions. This is a radical redesign of what storage solutions offer. With Qumulo, you can instantly understand: what files or applications are impacting capacity or performance. what data is hot, warm or cold. how 2015 data growth and performance trends inform your 2016 plans. RSVP Seating is limited. RSVP without delay to insure you see the latest solutions and talk directly to our product specialists. To RSVP, click here Where Le Ballpark - 6660 Rue Clark, Montral, QC H2S, Canada When Thursday, February 25th at 5:30pm We look forward to seeing you there! Link to Event:http://www.studiosysadmins.com/events/view/131/ -- Francis Provencher StudioSysAdmins Montreal Meetings Coordinator <francisp@gmail.com>

Flame 2016 + Z840 + Background reactor - PCoIP extenders?

February 25, 2016, 8:35 am

≫ Next: [Need Advice] Graphic Cards for Flame/Flare Workstation

≪ Previous: Reminder, this is Tomorrow: Montreal SSA Event: CEV/Qumulo Beer & Banter

$

0

0

Flame 2016 + Z840 + Background reactor - PCoIP extenders?

posted by Dan Young on Feb. 25, 2016, 11:35 a.m.

Hello everyone, With the advent of background reactor, what are you guys using for PCoIP? Do you have fully external extenders ala the Avocent HMX? Thinklogical? Have you sacrificed to remove something for the PCoIP card? Basically what I'm looking for, in my dream world, would be something that was capable of; - Completely external usage to CPU box, ala the HMX2050, no PCI slot card required (because Z840s are full in these configurations) - Capable of 2x DVI-D GUI panels, USB 2.0 or higher - Network extended (LAGs are OK), not fibre (willing to compromise on this if I can get exactly what I want for a good value over fibre) Does such a unicorn exist? Inb4 vendor email flood - seriously, do not flood me with spam from your company, vendors - this is your warning. Best, DY -- 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, With the advent of background reactor, what are you guys using for PCoIP? Do you have fully external extenders ala the Avocent HMX? Thinklogical? Have you sacrificed to remove something for the PCoIP card? Basically what I'm looking for, in my dream world, would be something that was capable of; - Completely external usage to CPU box, ala the HMX2050, no PCI slot card required (because Z840s are full in these configurations) - Capable of 2x DVI-D GUI panels, USB 2.0 or higher - Network extended (LAGs are OK), not fibre (willing to compromise on this if I can get exactly what I want for a good value over fibre) Does such a unicorn exist? Inb4 vendor email flood - seriously, do not flood me with spam from your company, vendors - this is your warning. Best, DY -- Dan Young Lead Systems Engineer LondonNew YorkLos AngelesMontral T+1 212 775 0600 135 Spring Street, New York NY 10012 TwitterFacebookframestore.com

[Need Advice] Graphic Cards for Flame/Flare Workstation

February 25, 2016, 12:35 pm

≫ Next: [Need Advice] Graphic Cards for Flame/Flare Workstation

≪ Previous: Flame 2016 + Z840 + Background reactor - PCoIP extenders?

$

0

0

[Need Advice] Graphic Cards for Flame/Flare Workstation

posted by David Kossin on Feb. 25, 2016, 3:35 p.m.

Hi Everyone! I just now discovered and joined this community. I am the sole IT person for a small VFX house so I hope I'm in the right place. I apologize in advance if I'm using this forum incorrectly.  I'm looking to upgrade the graphics cards in all of our primary workstations to help performance. I've chosen a card and have come up with a plan to improve all the systems and I would love to know if this is a sound idea.  We have 6 Z800/Z820 Flame workstations and 4 custom built Flare workstations. The Flames each have a Quadro 6000 card. The Flares each have a GeForce GTX 970 card in them. My plan is to purchase 6 Quadro M4000 cards to replace the Quadro 6000 cards in the Flames Then take the Quadro 6000 cards from the Flames and put them in the Flares. And finally, take all the left over cards and put them in boxes as either Burn nodes or add them to our RenderFarm for the CG guys.  I believe I have done proper research on these cards but it is still hard to tell if we will see significant improvements in performance on each of the machines. Does anyone have a similar experience they can share about doing a graphics card upgrade like this? I picked the Quadro M4000 because it seems to be the best cost to performance card. Let me know what you guys think of my plan.    Thanks!

Thread Tags:   workstation

0 Responses     0 Plus One's     0 Comments

 

Hi Everyone! I just now discovered and joined this community. I am the sole IT person for a small VFX house so I hope I'm in the right place. I apologize in advance if I'm using this forum incorrectly.  I'm looking to upgrade the graphics cards in all of our primary workstations to help performance. I've chosen a card and have come up with a plan to improve all the systems and I would love to know if this is a sound idea.  We have 6 Z800/Z820 Flame workstations and 4 custom built Flare workstations. The Flames each have a Quadro 6000 card. The Flares each have a GeForce GTX 970 card in them. My plan is to purchase 6 Quadro M4000 cards to replace the Quadro 6000 cards in the Flames Then take the Quadro 6000 cards from the Flames and put them in the Flares. And finally, take all the left over cards and put them in boxes as either Burn nodes or add them to our RenderFarm for the CG guys.  I believe I have done proper research on these cards but it is still hard to tell if we will see significant improvements in performance on each of the machines. Does anyone have a similar experience they can share about doing a graphics card upgrade like this? I picked the Quadro M4000 because it seems to be the best cost to performance card. Let me know what you guys think of my plan.    Thanks!

---

[Need Advice] Graphic Cards for Flame/Flare Workstation

February 25, 2016, 1:00 pm

≫ Next: Flame 2016 + Z840 + Background reactor - PCoIP extenders?

≪ Previous: [Need Advice] Graphic Cards for Flame/Flare Workstation

$

0

0

[Need Advice] Graphic Cards for Flame/Flare Workstation

posted by Saker Klippsten on Feb. 25, 2016, 4 p.m. (1 day ago)

Hi David. You should see the supported config for Flame. Its Quadro 6000 or M6000 for Z820/Z840 https://knowledge.autodesk.com/search-result/caas/sfdcarticles/sfdcarticles/flame-sysreqs.html For Z800. Quadro 6000. I strongly suggest sticking with these cards for your Z800's For the Z820/840 M6000. Flares you can use TitanX Cards which is similar config to the m6000 pretty much. @ 1000$ each. -S On Thu, Feb 25, 2016 at 3:35 PM, David Kossin <content@studiosysadmins.com> wrote: Hi Everyone! I just now discovered and joined this community. I am the sole IT person for a small VFX house so I hope I'm in the right place. I apologize in advance if I'm using this forum incorrectly. I'm looking to upgrade the graphics cards in all of our primary workstations to help performance. I've chosen a card and have come up with a plan to improve all the systems and I would love to know if this is a sound idea. We have 6 Z800/Z820 Flame workstations and 4 custom built Flare workstations. The Flames each have a Quadro 6000 card. The Flares each have aGeForce GTX 970 card in them. My plan is to purchase 6Quadro M4000 cards to replace the Quadro 6000 cards in the Flames Then take the Quadro 6000 cards from the Flames and put them in the Flares. And finally, take all the left over cards and put them in boxes as either Burn nodes or add them to our RenderFarm for the CG guys. I believe I have done proper research on these cards but it is still hard to tell if we will see significant improvements in performance on each of the machines. Does anyone have a similar experience they can share about doing a graphics card upgrade like this? I picked the Quadro M4000 because it seems to be the best cost to performance card. Let me know what you guys think of my plan. Thanks! 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

 

Hi David. You should see the supported config for Flame. Its Quadro 6000 or M6000 for Z820/Z840 https://knowledge.autodesk.com/search-result/caas/sfdcarticles/sfdcarticles/flame-sysreqs.html For Z800. Quadro 6000. I strongly suggest sticking with these cards for your Z800's For the Z820/840 M6000. Flares you can use TitanX Cards which is similar config to the m6000 pretty much. @ 1000$ each. -S On Thu, Feb 25, 2016 at 3:35 PM, David Kossin <content@studiosysadmins.com> wrote: Hi Everyone! I just now discovered and joined this community. I am the sole IT person for a small VFX house so I hope I'm in the right place. I apologize in advance if I'm using this forum incorrectly. I'm looking to upgrade the graphics cards in all of our primary workstations to help performance. I've chosen a card and have come up with a plan to improve all the systems and I would love to know if this is a sound idea. We have 6 Z800/Z820 Flame workstations and 4 custom built Flare workstations. The Flames each have a Quadro 6000 card. The Flares each have aGeForce GTX 970 card in them. My plan is to purchase 6Quadro M4000 cards to replace the Quadro 6000 cards in the Flames Then take the Quadro 6000 cards from the Flames and put them in the Flares. And finally, take all the left over cards and put them in boxes as either Burn nodes or add them to our RenderFarm for the CG guys. I believe I have done proper research on these cards but it is still hard to tell if we will see significant improvements in performance on each of the machines. Does anyone have a similar experience they can share about doing a graphics card upgrade like this? I picked the Quadro M4000 because it seems to be the best cost to performance card. Let me know what you guys think of my plan. Thanks! To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe

Flame 2016 + Z840 + Background reactor - PCoIP extenders?

February 25, 2016, 3:00 pm

≫ Next: [Need Advice] Graphic Cards for Flame/Flare Workstation

≪ Previous: [Need Advice] Graphic Cards for Flame/Flare Workstation

$

0

0

Flame 2016 + Z840 + Background reactor - PCoIP extenders?

posted by Peter Smith on Feb. 25, 2016, 6 p.m. (1 day ago)

Hola! Adder. http://www.adder.com/products/adderlink-infinity-dual Hit your Flame with the "no dithering" fix, and you're golden*. Bonus, you won't need to rob a bank. *except old Wacoms and various other things. Not totally golden. Still pretty shiny. HTH Pete p.s. Adder have new stuff** which allows moar pixels. ** Ain't seen it. YMMV. -- On 25/02/16 16:29, Dan Young wrote: Hello everyone, With the advent of background reactor, what are you guys using for PCoIP? Do you have fully external extenders ala the Avocent HMX? Thinklogical? Have you sacrificed to remove something for the PCoIP card? Basically what I'm looking for, in my dream world, would be something that was capable of; - Completely external usage to CPU box, ala the HMX2050, no PCI slot card required (because Z840s are full in these configurations) - Capable of 2x DVI-D GUI panels, USB 2.0 or higher - Network extended (LAGs are OK), not fibre (willing to compromise on this if I can get exactly what I want for a good value over fibre) Does such a unicorn exist? Inb4 vendor email flood - seriously, do not flood me with spam from your company, vendors - this is your warning. Best, DY -- 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

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

Hola! Adder. http://www.adder.com/products/adderlink-infinity-dual Hit your Flame with the "no dithering" fix, and you're golden*. Bonus, you won't need to rob a bank. *except old Wacoms and various other things. Not totally golden. Still pretty shiny. HTH Pete p.s. Adder have new stuff** which allows moar pixels. ** Ain't seen it. YMMV. -- On 25/02/16 16:29, Dan Young wrote: Hello everyone, With the advent of background reactor, what are you guys using for PCoIP? Do you have fully external extenders ala the Avocent HMX? Thinklogical? Have you sacrificed to remove something for the PCoIP card? Basically what I'm looking for, in my dream world, would be something that was capable of; - Completely external usage to CPU box, ala the HMX2050, no PCI slot card required (because Z840s are full in these configurations) - Capable of 2x DVI-D GUI panels, USB 2.0 or higher - Network extended (LAGs are OK), not fibre (willing to compromise on this if I can get exactly what I want for a good value over fibre) Does such a unicorn exist? Inb4 vendor email flood - seriously, do not flood me with spam from your company, vendors - this is your warning. Best, DY -- 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

[Need Advice] Graphic Cards for Flame/Flare Workstation

February 26, 2016, 4:25 am

≫ Next: Web scale data growth

≪ Previous: Flame 2016 + Z840 + Background reactor - PCoIP extenders?

$

0

0

[Need Advice] Graphic Cards for Flame/Flare Workstation

posted by  on Feb. 26, 2016, 7:25 a.m.

We tried Flame with an M4000 recently and were pretty pleased with the results, given that we are running Quadro 6000 and in some cases even FX 5800 cards.  I'm not sure if it's still the case, but the DKU at the time contained an NVIDIA driver version that was just slightly too old for the M4000 and M5000 cards.  I mention this only because there is a slight source code modification (one line, IIRC) on the drivers distributed in the DKU -- we chose to replicate this change in the nearest driver with M4000 support (346.87). Cheers, Tom @ Absolute From: "David Kossin" <content@studiosysadmins.com> To: studiosysadmins-discuss@studiosysadmins.com Sent: Thursday, 25 February, 2016 8:35:29 PM Subject: [SSA-Discuss] [Need Advice] Graphic Cards for Flame/Flare        Workstation Hi Everyone! I just now discovered and joined this community. I am the sole IT person for a small VFX house so I hope I'm in the right place. I apologize in advance if I'm using this forum incorrectly.  I'm looking to upgrade the graphics cards in all of our primary workstations to help performance. I've chosen a card and have come up with a plan to improve all the systems and I would love to know if this is a sound idea.  We have 6 Z800/Z820 Flame workstations and 4 custom built Flare workstations. The Flames each have a Quadro 6000 card. The Flares each have a GeForce GTX 970 card in them. My plan is to purchase 6 Quadro M4000 cards to replace the Quadro 6000 cards in the Flames Then take the Quadro 6000 cards from the Flames and put them in the Flares. And finally, take all the left over cards and put them in boxes as either Burn nodes or add them to our RenderFarm for the CG guys.  I believe I have done proper research on these cards but it is still hard to tell if we will see significant improvements in performance on each of the machines. Does anyone have a similar experience they can share about doing a graphics card upgrade like this? I picked the Quadro M4000 because it seems to be the best cost to performance card. Let me know what you guys think of my plan.    Thanks! 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

 

We tried Flame with an M4000 recently and were pretty pleased with the results, given that we are running Quadro 6000 and in some cases even FX 5800 cards.  I'm not sure if it's still the case, but the DKU at the time contained an NVIDIA driver version that was just slightly too old for the M4000 and M5000 cards.  I mention this only because there is a slight source code modification (one line, IIRC) on the drivers distributed in the DKU -- we chose to replicate this change in the nearest driver with M4000 support (346.87). Cheers, Tom @ Absolute From: "David Kossin" <content@studiosysadmins.com> To: studiosysadmins-discuss@studiosysadmins.com Sent: Thursday, 25 February, 2016 8:35:29 PM Subject: [SSA-Discuss] [Need Advice] Graphic Cards for Flame/Flare        Workstation Hi Everyone! I just now discovered and joined this community. I am the sole IT person for a small VFX house so I hope I'm in the right place. I apologize in advance if I'm using this forum incorrectly.  I'm looking to upgrade the graphics cards in all of our primary workstations to help performance. I've chosen a card and have come up with a plan to improve all the systems and I would love to know if this is a sound idea.  We have 6 Z800/Z820 Flame workstations and 4 custom built Flare workstations. The Flames each have a Quadro 6000 card. The Flares each have a GeForce GTX 970 card in them. My plan is to purchase 6 Quadro M4000 cards to replace the Quadro 6000 cards in the Flames Then take the Quadro 6000 cards from the Flames and put them in the Flares. And finally, take all the left over cards and put them in boxes as either Burn nodes or add them to our RenderFarm for the CG guys.  I believe I have done proper research on these cards but it is still hard to tell if we will see significant improvements in performance on each of the machines. Does anyone have a similar experience they can share about doing a graphics card upgrade like this? I picked the Quadro M4000 because it seems to be the best cost to performance card. Let me know what you guys think of my plan.    Thanks! To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe

Web scale data growth

February 26, 2016, 11:20 am

≫ Next: SMB 10Gb switch

≪ Previous: [Need Advice] Graphic Cards for Flame/Flare Workstation

$

0

0

Web scale data growth

posted by Jean-Francois Panisset on Feb. 26, 2016, 2:20 p.m.

Google has published a paper about what they would like to see in future hard drives: http://static.googleusercontent.com/media/research.google.com/en//pubs/archive/44830.pdf One interesting tidbit: YouTube data growth is currently at 1PB per day. And you thought you had trouble managing cache data from your FX team... JF

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

Google has published a paper about what they would like to see in future hard drives: http://static.googleusercontent.com/media/research.google.com/en//pubs/archive/44830.pdf One interesting tidbit: YouTube data growth is currently at 1PB per day. And you thought you had trouble managing cache data from your FX team... JF

SMB 10Gb switch

March 1, 2016, 7:10 am

≫ Next: maya on mac, questions

≪ Previous: Web scale data growth

$

0

0

SMB 10Gb switch

posted by Doug Meyer on March 1, 2016, 10:10 a.m. (3 days ago)

So I'm searching for a relatively low cost 10Gb switch for a small studio application with the following requirements: 24 10Gb copper + 2 or more SFP uplinks able to stack 2 for failover link aggregation reasonable management capabilities POE unnecessary Ran across the Cisco SG350XG-24T  http://www.provantage.com/cisco-systems-sg350xg-24t-k9~7CSSM06Q.htm There is also a Cisco SG550XG-24T but the only difference i can tell is it allows stacking 8 vs 4 units Are there any better suggestions sub $5K ? thnx

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

So I'm searching for a relatively low cost 10Gb switch for a small studio application with the following requirements: 24 10Gb copper + 2 or more SFP uplinks able to stack 2 for failover link aggregation reasonable management capabilities POE unnecessary Ran across the Cisco SG350XG-24T  http://www.provantage.com/cisco-systems-sg350xg-24t-k9~7CSSM06Q.htm There is also a Cisco SG550XG-24T but the only difference i can tell is it allows stacking 8 vs 4 units Are there any better suggestions sub $5K ? thnx

maya on mac, questions

March 1, 2016, 6:25 pm

≫ Next: ffmpeg linear exr to sRGB/rec709 quicktime

≪ Previous: SMB 10Gb switch

$

0

0

maya on mac, questions

posted by Ben De Luca on March 1, 2016, 9:25 p.m. (3 days ago)

Hey Guys, I am doing some consulting work and I am trying to make maya work on my mac laptop, I wondered 1. has any one got it running on el capitan? (I stupidly upgraded) It bombs out on settings related to the new colour management stuff. 2. has any one been able to download maya 2105 sp6 for mac, every time I try it is corrupted. I cant tell if its my terrible internet or the upload is broken. I am working through to autodesk with my clients IT but I thought I should ask the collective. -Ben

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

Hey Guys, I am doing some consulting work and I am trying to make maya work on my mac laptop, I wondered 1. has any one got it running on el capitan? (I stupidly upgraded) It bombs out on settings related to the new colour management stuff. 2. has any one been able to download maya 2105 sp6 for mac, every time I try it is corrupted. I cant tell if its my terrible internet or the upload is broken. I am working through to autodesk with my clients IT but I thought I should ask the collective. -Ben

ffmpeg linear exr to sRGB/rec709 quicktime

March 2, 2016, 3:20 am

≫ Next: [ANN] MacDevOps:YVR registration open

≪ Previous: maya on mac, questions

$

0

0

ffmpeg linear exr to sRGB/rec709 quicktime

posted by Erik Johansson on March 2, 2016, 6:20 a.m. (2 days ago)

I figured someone here might possibly have setup some batch encoding of exr to quicktime files with ffmpeg. Could you get colorspace conversions working nicely? How did you do it? I am struggling finding the right ffmpeg flags if any exists. Cheers, Erik

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

I figured someone here might possibly have setup some batch encoding of exr to quicktime files with ffmpeg. Could you get colorspace conversions working nicely? How did you do it? I am struggling finding the right ffmpeg flags if any exists. Cheers, Erik

---

[ANN] MacDevOps:YVR registration open

March 2, 2016, 5:05 am

≫ Next: Obscure Windows DFS Hyper-V nonsense

≪ Previous: ffmpeg linear exr to sRGB/rec709 quicktime

$

0

0

[ANN] MacDevOps:YVR registration open

posted by Mathieu Mauser on March 2, 2016, 8:05 a.m. (2 days ago)

FYI: Just wanted to let you know that MacDevOps:YVR registration is now open. We had some studios attend last year, and there are so many more in Vancouver now. (If theres interest perhaps next year Ill organize one of the days for VFX DevOps. Sound cool?). MacDevOps will be held on June 20-21, 2016 at SFU Harbour Centre in downtown Vancouver. We have awesome international speakers lined up for this years 2-day conference. Topics include: Ansible, Django, Puppet, Docker, Jenkins, Munki, Git, etc Check out the speakers page: http://www.macdevops.ca/speakers/ Many more to be announced soon. MacDevOps:YVR 2016 registration is $399 CDN Early Bird. (Price goes up to $499 CDN on April 2nd.) For the reg link and more information please visit our website: http://www.macdevops.ca/ :) Mat X @macdevopsYVR 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

 

FYI: Just wanted to let you know that MacDevOps:YVR registration is now open. We had some studios attend last year, and there are so many more in Vancouver now. (If theres interest perhaps next year Ill organize one of the days for VFX DevOps. Sound cool?). MacDevOps will be held on June 20-21, 2016 at SFU Harbour Centre in downtown Vancouver. We have awesome international speakers lined up for this years 2-day conference. Topics include: Ansible, Django, Puppet, Docker, Jenkins, Munki, Git, etc Check out the speakers page: http://www.macdevops.ca/speakers/ Many more to be announced soon. MacDevOps:YVR 2016 registration is $399 CDN Early Bird. (Price goes up to $499 CDN on April 2nd.) For the reg link and more information please visit our website: http://www.macdevops.ca/ :) Mat X @macdevopsYVR To unsubscribe from the list send a blank e-mail to mailto:studiosysadmins-discuss-request@studiosysadmins.com?subject=unsubscribe

Obscure Windows DFS Hyper-V nonsense

March 3, 2016, 11:25 am

≫ Next: 3ds Max slowness working off of a network vs locally

≪ Previous: [ANN] MacDevOps:YVR registration open

$

0

0

Obscure Windows DFS Hyper-V nonsense

posted by Jean-Francois Panisset on March 3, 2016, 2:25 p.m. (1 day ago)

Most likely a long shot, but I figured I might ask. I have a Windows 2012R2 VM on a Hyper-V cluster where DFS seems to be confused as to which AD site the machine is in: it seems that DFS is using the 169.254.x.x IP from the "Failover Cluster Virtual Adapter" to figure out what site it is in, and of course that network doesn't exist as an AD entity, so "dfsutil /sitename:my_dc" returns "no site association". I've played around with techniques to re-order network interfaces, and the main interface is now showing up first, but DFS insists on using 169.254.x.x to figure out which site it's in. Other machines see that server in the right AD site. Anyone run into something like that? JF

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     1 Comments

 

Most likely a long shot, but I figured I might ask. I have a Windows 2012R2 VM on a Hyper-V cluster where DFS seems to be confused as to which AD site the machine is in: it seems that DFS is using the 169.254.x.x IP from the "Failover Cluster Virtual Adapter" to figure out what site it is in, and of course that network doesn't exist as an AD entity, so "dfsutil /sitename:my_dc" returns "no site association". I've played around with techniques to re-order network interfaces, and the main interface is now showing up first, but DFS insists on using 169.254.x.x to figure out which site it's in. Other machines see that server in the right AD site. Anyone run into something like that? JF

3ds Max slowness working off of a network vs locally

March 3, 2016, 4:34 pm

≫ Next: Cloud rendering

≪ Previous: Obscure Windows DFS Hyper-V nonsense

$

0

0

3ds Max slowness working off of a network vs locally

posted by SeanFarrell on March 3, 2016, 7:34 p.m. (1 day ago)

Hey everyone!   Our dynamics artists have been running into a problem using 3ds max where the program is very slow to render and read/write cahes to a network location.  Copying files up and down to the same network works as expected and utilizes all of the bandwith available, but the max process only uses a small fraction of what is available resulting in painfully slow processing.  When running the same process locally, speeds are about five times as fast.  Has anyone else experineced such strangeness?  We are running vray 3.3 for rendering and 3ds max 2016.   Thanks!   Sean

Thread Tags:   network 3ds autodesk, max, vray, renders,

0 Responses     0 Plus One's     0 Comments

 

Hey everyone!   Our dynamics artists have been running into a problem using 3ds max where the program is very slow to render and read/write cahes to a network location.  Copying files up and down to the same network works as expected and utilizes all of the bandwith available, but the max process only uses a small fraction of what is available resulting in painfully slow processing.  When running the same process locally, speeds are about five times as fast.  Has anyone else experineced such strangeness?  We are running vray 3.3 for rendering and 3ds max 2016.   Thanks!   Sean

Cloud rendering

March 4, 2016, 1:53 am

≫ Next: How do you present data from multiple storage sources to your users?

≪ Previous: 3ds Max slowness working off of a network vs locally

$

0

0

Cloud rendering

posted by Doug Mardon on March 4, 2016, 4:53 a.m.

Hello all, Assuming this is the correct place to ask (apologies in advance if not), I have a question directed primarily at VFX Studios based in London, UK; however input from those of you based elsewhere in the World would be greatfully recieved as well. I wanted to know: How often you outource your rendering requirments to cloud based providers such as Google CE or Amazon EC2? How much cost and/ or processing speed is a factor in this decision? And if you do outsource; what the frequency you use these services is? (1 month blocks or by the hour...) Many thanks in advance all! Doug

Thread Tags:   render

0 Responses     0 Plus One's     0 Comments

 

Hello all, Assuming this is the correct place to ask (apologies in advance if not), I have a question directed primarily at VFX Studios based in London, UK; however input from those of you based elsewhere in the World would be greatfully recieved as well. I wanted to know: How often you outource your rendering requirments to cloud based providers such as Google CE or Amazon EC2? How much cost and/ or processing speed is a factor in this decision? And if you do outsource; what the frequency you use these services is? (1 month blocks or by the hour...) Many thanks in advance all! Doug

How do you present data from multiple storage sources to your users?

March 4, 2016, 9:45 am

≫ Next: How do you present data from multiple storage sources to your users?

≪ Previous: Cloud rendering

$

0

0

How do you present data from multiple storage sources to your users?

posted by Rory Falloon on March 4, 2016, 12:45 p.m.

Hi, Once upon a time, using the BlueArc for global symlinks took a headache out of presenting files from different servers (isilon, bluearc, supermicro nearline) as being from one source to both linux AND windows users. How are people doing it these days? DFS? Another way? Interested to hear. Rory.

Thread Tags:   discuss-at-studiosysadmins

0 Responses     0 Plus One's     0 Comments

 

Hi, Once upon a time, using the BlueArc for global symlinks took a headache out of presenting files from different servers (isilon, bluearc, supermicro nearline) as being from one source to both linux AND windows users. How are people doing it these days? DFS? Another way? Interested to hear. Rory.