Re: [j-nsp] logical system in production - MX960
> Of Saku Ytti > Sent: Tuesday, October 03, 2017 9:26 AM > > Since node slicing, I think logical-systems are rather useless. > > https://www.juniper.net/documentation/en_US/junos/topics/concept/nod > e-slicing-overview.html > True that, Though it will take couple years till the whole circus becomes production ready, I very much welcome this external CP approach especially for big chassis like MX960 where these could become full with just couple of FPCs installed in the box just because CP can't cope with the combination of BGP sessions, VRFs and prefixes (sad story). adam ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
Re: [j-nsp] logical system in production - MX960
Well, I have heard that the external server is for scaling node slicing beyond what the stand-alone node slicing in-chassis can do. Would someone have a link or instructions on how to do node slicing *without* the external x86 server ? -Aaron ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
Re: [j-nsp] logical system in production - MX960
Strongly agree with Ytti. Most LSYS deployments I've seen were really "deploy an LSYS to deploy an LSYS" kind of of thing. Some more or less viable use-case I know is BGP scaling: dedicated rpd in an LSYS for an off-path RR. It creates a dedicated 32-bit rpd, which can consume another 4GB of RAM. But nowadays, with all the kinds of VM-based routers/RRs, 64-bit rpd, node slicing and stuff, this approach looks to be obsolete. 2017-10-03 10:25 GMT+02:00 Saku Ytti: > On 25 September 2017 at 15:10, Aaron Gould wrote: > > > Hey, > > > Do you all use logical systems in your production environment ? > > No. Be sure you're solving real technical problems, not political or > process. I know there are use-cases, but I suspect there are fewer > use-cases than deployments. > > There are several layers of function multiplexting > > virtual-router: essentially just VRF > logical-system: separate rpd running in same control-plane > node slicing: separate freebsd KVM running on single hypervisor > > Since node slicing, I think logical-systems are rather useless. > > https://www.juniper.net/documentation/en_US/junos/ > topics/concept/node-slicing-overview.html > > -- > ++ytti > ___ > juniper-nsp mailing list juniper-nsp@puck.nether.net > https://puck.nether.net/mailman/listinfo/juniper-nsp > ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
Re: [j-nsp] logical system in production - MX960
Thanks, and I'm learning that node slicing requires at least an MX480...also supported on MX960, MX2010, MX2020node slicing is very different than lsys in that it runs a few of its components on external x86 servers... external meaning separate from the bsys (bsys is the MX hardware chassis). Components... https://www.juniper.net/documentation/en_US/junos/topics/concept/node-slicing-understanding.html feature guide...84 pages https://www.juniper.net/documentation/en_US/junos/information-products/pathway-pages/junos-node-slicing/junos-node-slicing.pdf - Aaron Gould ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
Re: [j-nsp] logical system in production - MX960
On 25 September 2017 at 15:10, Aaron Gouldwrote: Hey, > Do you all use logical systems in your production environment ? No. Be sure you're solving real technical problems, not political or process. I know there are use-cases, but I suspect there are fewer use-cases than deployments. There are several layers of function multiplexting virtual-router: essentially just VRF logical-system: separate rpd running in same control-plane node slicing: separate freebsd KVM running on single hypervisor Since node slicing, I think logical-systems are rather useless. https://www.juniper.net/documentation/en_US/junos/topics/concept/node-slicing-overview.html -- ++ytti ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
Re: [j-nsp] logical system in production - MX960
> Of Aaron Gould > Sent: Monday, September 25, 2017 1:11 PM > > Do you all use logical systems in your production environment ? > No I'm not suicidal, but then again if you test all aspects and it works for your setup what is there to object to right? Good luck. adam ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
Re: [j-nsp] logical system in production - MX960
On Mon, Sep 25, 2017 at 07:10:47AM -0500, Aaron Gould wrote: > A few questions about logical systems. related to a new 5-node MX960 100 gig > ring. > > Do you all use logical systems in your production environment ? I do. > Do you contain your core P functions inside of an lsys ? My network is rather small and combines P/PE functions, so I guess the answer is "yes". > Is the master/hosting lsys completely absent of any and all provider routes, > and does it only contain a simply management interface (fxp) and not much > more than that ? No, I use the main systems to run my LAN and a single WAN lsys on each MX480. They are interconnected by my firewalls. > If you were deploying a 5-node MX960 mpls core for an ISP, would you deploy > lsys's from the start and put the core routing inside an lsys ? Maybe, but there are some limitations. Some things don't work in lsys, such as port-mirroring & inline-jflow sampling (although support for the latter is supposed to be there in the latest Junos releases--it was buggy/crashy and I had to turn it off). https://www.juniper.net/documentation/en_US/junos/topics/concept/logical-systems-restrictions.html ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
[j-nsp] logical system in production - MX960
A few questions about logical systems. related to a new 5-node MX960 100 gig ring. Do you all use logical systems in your production environment ? Do you contain your core P functions inside of an lsys ? Is the master/hosting lsys completely absent of any and all provider routes, and does it only contain a simply management interface (fxp) and not much more than that ? If you were deploying a 5-node MX960 mpls core for an ISP, would you deploy lsys's from the start and put the core routing inside an lsys ? I use lsys in my lab mx104 and have all p/pe/ce/c functions inside lots of different lsys's.. but now I want to know how everyone uses lsys for production as mentioned above. -Aaron Gould ___ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp