Hey Sarin Could you please provide me the status of the bugs we filed? (Contrail Command UI) 1. https://bugs.launchpad.net/juniperopenstack/+bug/1791842 1. https://bugs.launchpad.net/juniperopenstack/+bug/1792650 Thanks, Ashutosh From: Sarin Kizhakkepurayil Date: Friday, September 14, 2018 at 11:00 AM To: Ashutosh Gupta , Radosław Romaszewski Cc: Dilip Sundarraj Subject: Re: 5.0.01 /Try Upgrade and Documentation - Update on Contrail Command Issues +Radek, please take a look at the issue below Ashutosh, could you open up a bug and assign to Radek. He will get it fixed. Thanks From: Ashutosh Gupta Date: Friday, September 14, 2018 at 10:01 AM To: Sarin Kizhakkepurayil Cc: Dilip Sundarraj Subject: FW: 5.0.01 /Try Upgrade and Documentation - Update on Contrail Command Issues Hello Sarin Can you please take a look at the issue in the trailing mail. Thanks, Ashutosh From: Linh Phan Date: Thursday, September 13, 2018 at 12:02 PM To: Ali Zaringhalam , Ashutosh Gupta , Chaitanya Kadiyala Cc: Dilip Sundarraj Subject: RE: 5.0.01 /Try Upgrade and Documentation - Update on Contrail Command Issues Hi Ashutosh, I wonder if this is an expected behavior of CC For the vSRX firewalls as “Service_chaining”, under tab Instances, it shows the IP address in this strange format when I hover my mouse to both firewall instance ? [cid:image004.png@01D44B59.57790AF0] [cid:image005.png@01D44B59.57790AF0] And the only place under Instances tab that I can look for the correct IP address (to SSH to it) is under the json snippet [cid:image006.png@01D44B59.57790AF0] Meanwhile in Horizon it shows the IPs as normal [cid:image007.png@01D44B59.57790AF0] [cid:image001.png@01D44B59.9ACC4D80] Only the GUI display have problem. All instances are running and networking correctly. Thanks. Regards, Linh. From: Ali Zaringhalam Sent: Wednesday, September 12, 2018 7:10 PM To: Ashutosh Gupta ; Linh Phan ; Chaitanya Kadiyala Cc: Dilip Sundarraj Subject: RE: 5.0.01 /Try Upgrade and Documentation - Update on Contrail Command Issues Thanks Ashutosh. I am following progress on related threads. Also, as discussed today, we look forward to information on how to pass Cassandra parameters (e,g., heap size) to the Cassandra containers before they are started. Regards, Ali From: Ashutosh Gupta Sent: Wednesday, September 12, 2018 8:00 PM To: Ali Zaringhalam <>>; Linh Phan <>>; Chaitanya Kadiyala <>> Cc: Dilip Sundarraj <>> Subject: FW: 5.0.01 /Try Upgrade and Documentation - Update on Contrail Command Issues Hello Ali/Chaitanya/Linh I just had a talk with Sarin and this is what he said for the following issues 1. Contrail Service Chain inactive alarm on Contrail Command GUI. Bug Filed https://bugs.launchpad.net/juniperopenstack/+bug/1791842 (Might need a week or so to resolve) 1. This VNC params in nova.conf is not taking effects. This makes it not possible to directly access instance’s VNC console, from both Horizon/ContrailCommand. Ignitious is having a look at this issue. I should be able to update you on this by tomorrow EOD. From: Ashutosh Gupta <>> Date: Monday, September 10, 2018 at 10:32 AM To: Ali Zaringhalam <>>, Chaitanya Kadiyala <>>, Linh Phan <>> Cc: Dilip Sundarraj <>> Subject: Re: 5.0.01 /Try Upgrade and Documentation - Working Session Hello Ali/Chaitanya/Linh As per our conclusions from the meeting, these the issues we are left to address. 1. Contrail Service Chain inactive alarm on Contrail Command GUI. 2. This VNC params in nova.conf is not taking effects. This makes it not possible to directly access instance’s VNC console, from both Horizon/ContrailCommand. As discussed, I will provide the setup access to Sarin and Ignatious and they will look into the issue. I will follow up with him thereafter. Thanks, Ashutosh From: Ali Zaringhalam <>> Date: Monday, September 10, 2018 at 8:59 AM To: Ashutosh Gupta <>>, Chaitanya Kadiyala <>>, Linh Phan <>> Cc: Dilip Sundarraj <>> Subject: RE: 5.0.01 /Try Upgrade and Documentation - Working Session Hi Ashutosh: Please add the Service Chain issue Linh (see his email) identified for Contrail Command. Regards, Ali From: Ashutosh Gupta Sent: Monday, September 10, 2018 11:35 AM To: Chaitanya Kadiyala <>>; Ali Zaringhalam <>>; Linh Phan <>> Cc: Dilip Sundarraj <>> Subject: Re: 5.0.01 /Try Upgrade and Documentation - Working Session Hello Chaitanya/Ali/Linh Sorry for the delayed response. List of Issues 1. How do I change/increase the timeout for Horizon/Contrail GUI ? Many users have complained about the GUI keep timing out while they’re doing demo. UI works based on keystone token expiry. Extending default expiry on keystone should work (UI team has not tested this) 1. Is there an easy way to add customized “nova.conf” params when installing with Contrail Command ? For example, if manually install with ansible-deployer, the instances.yaml have this section. The only way I can do it now is by manually add those line in all nova.conf kolla_config: openstack_release: ocata customize: nova.conf: | [oslo_messaging_notifications] driver = messagingv2 [DEFAULT] disk_allocation_ratio = 20.0 ram_allocation_ratio = 5.0 cpu_allocation_ratio = 25.0 From UI setup wizard, under advanced options in orchestrator step, you could set kolla globals key value params. but I believe sending second level key value will not work (Ram/Ignatious still need to confirm this) 1. This VNC params in nova.conf is not taking effects. This makes it not possible to directly access instance’s VNC console, from both Horizon/ContrailCommand. Under Investigation from the engineering team. 1. Memory: with 64 G RAM per cfg/ctlr/analytics node, we expect the number of PODs per server to be reduced to 2-3 from the current 5 PODs/server in Contrail 3.2.3. This is a significant reduction and will impact the intended applications (external users and training). We have asked engineering to investigate tuning parameters to reduce the RAM utilization. The difference is micro-services. 32 GB is just about enough to bring up the container. Whether we can fine tune, I still need a confirmation on this from the development team. 1. Contrail Command: when we install Contrail Command, we configure the cfg node for CSN. This causes vRouter to be installed on the cfg node. We currently don’t plan to use CSN but would like to make sure that the presence of vrouter on cfg node is benign. Should we remove CSN from the installation altogether? vRouter is usually not present on the config node. CSN is not necessary for bringing up the cluster and you can remove CSN from the installation. 1. “Contrail Command resources: what is the minimum requirement for a VM hosting Contrail Command. In our setup, each POD is managed by a separate Contrail Command. We have installed this on a VM with 2 CPU and 2 G RAM and it seems to working fine. We would like to have confirmation that this is OK.“ 2G of RAM is not tested and could be too less. The min recommended is 8GB and 2 CPU. Linh Phan is testing with 4BG to see if it is required to be upgraded to 8GB. I will confirm the config node RAM issue with the engineering team today. Thanks, Ashutosh From: Dilip Sundarraj <>> on behalf of Chaitanya Kadiyala <>> Date: Saturday, September 8, 2018 at 7:48 AM To: Ashutosh Gupta <>> Subject: FW: 5.0.01 /Try Upgrade and Documentation - Working Session Hi Ashutosh Please provide an update of where things stand and what the open issues are. -Dilip From: > When: 10:00 AM - 11:00 AM September 10, 2018 Subject: 5.0.01 /Try Upgrade and Documentation - Working Session Location: ~CR-US-SVL-A.6.103-Branch Rickey; Skype Hi Ashutosh, +Dilip, As we discussed on the call, can you talk to Sarin and other folks from engineering who can help with the issues related to Contrail Command and CSN. I will setup a call for Monday between all the people listed on this thread to go over the details. Thanks Chaitanya ......................................................................................................................................... Join Skype Meeting Trouble Joining? Try Skype Web App Join by phone +14089369000 (USA, West Coast) English (United States) +19785898300 (USA, West Coast) English (United States) ILYNC (45962) (USA, West Coast) English (United States) +18446454398 (USA, West Coast) English (United States) +18332135669 (USA, West Coast) English (United States) Find a local number Conference ID: 66533016 Forgot your dial-in PIN? |Help Please consider with whom you are communicating, including non-Juniper personnel of companies federated on Skype for Business, before sharing any confidential information. Non-disclosure agreements may apply. [!OC([1033])!] ......................................................................................................................................... _____________________________________________ From: Ali Zaringhalam Sent: Friday, September 7, 2018 10:34 AM To: Ashutosh Gupta <>>; Linh Phan <>> Cc: Chaitanya Kadiyala <>>; Sarin Kizhakkepurayil <>>; Dilip Sundarraj <>>; Ignatious Johnson <>>; Ramprakash Ram Mohan <>>; Jerrod Spooner <>> Subject: RE: 5.0.01 /Try Upgrade and Documentation - Working Session Hi Folks: Below are the meeting minutes from the conference call today. * Memory: with 64 G RAM per cfg/ctlr/analytics node, we expect the number of PODs per server to be reduced to 2-3 from the current 5 PODs/server in Contrail 3.2.3. This is a significant reduction and will impact the intended applications (external users and training). We have asked engineering to investigate tuning parameters to reduce the RAM utilization. I am attaching the recent email thread on this topic. * Contrail Command: when we install Contrail Command, we configure the cfg node for CSN. This causes vRouter to be installed on the cfg node. We currently don’t plan to use CSN but would like to make sure that the presence of vrouter on cfg node is benign. Should we remove CSN from the installation altogether? * Console access: the following thread discussed the issue. We are awaiting a resolution from engineering. * Contrail Command resources: what is the minimum requirement for a VM hosting Contrail Command. In our setup, each POD is managed by a separate Contrail Command. We have installed this on a VM with 2 CPU and 2 G RAM and it seems to working fine. We would like to have confirmation that this is OK. * Chaitanya will set up a follow-up call early next week. Regards, Ali From: Ashutosh Gupta Sent: Friday, September 7, 2018 12:34 PM To: Linh Phan <>> Cc: Ali Zaringhalam <>>; Chaitanya Kadiyala <>>; Sarin Kizhakkepurayil <>>; Dilip Sundarraj <>>; Ignatious Johnson <>>; Ramprakash Ram Mohan <>> Subject: FW: 5.0.01 /Try Upgrade and Documentation - Working Session +Linh, Ali, Chaitanya From: Sarin Kizhakkepurayil <>> Date: Friday, September 7, 2018 at 9:10 AM To: Ashutosh Gupta <>>, Ignatious Johnson <>>, Ramprakash Ram Mohan <>> Cc: Dilip Sundarraj <>> Subject: Re: 5.0.01 /Try Upgrade and Documentation - Working Session +Ignatious, Ram 1. UI works based on keystone token expiry. Extending default expiry on keystone should work (UI team has not tested this) 2. From UI setup wizard, under advanced options in orchestrator step, you could set kolla globals key value params. but I believe sending second level key value will not work. Ram/Ignatious can confirm this. 3. This needs some more investigation. We ask nova for the console URL and redirect user to same. There may be gap in base URL host:port as we’ve not tested with changing that. I’ll get back to you on this Thanks From: Ashutosh Gupta <>> Date: Friday, September 7, 2018 at 8:47 AM To: Sarin Kizhakkepurayil <>> Cc: Dilip Sundarraj <>> Subject: FW: 5.0.01 /Try Upgrade and Documentation - Working Session Hello Sarin Marketing team has some queries regarding Contrail Command Setup which is mentioned in the trailing email. Can you please answer some of the queries? Thanks Ashutosh From: Linh Phan <>> Date: Thursday, September 6, 2018 at 9:48 AM To: Ali Zaringhalam <>>, Ashutosh Gupta <>>, Chaitanya Kadiyala <>> Cc: Dilip Sundarraj <>> Subject: RE: 5.0.01 /Try Upgrade and Documentation - Working Session Hi Ashutosh, I have some questions regards to Contrail Command setup. 1. How do I change/increase the timeout for Horizon/Contrail GUI ? Many users have complained about the GUI keep timing out while they’re doing demo. 1. Is there an easy way to add customized “nova.conf” params when installing with Contrail Command ? For example, if manually install with ansible-deployer, the instances.yaml have this section. The only way I can do it now is by manually add those line in all nova.conf kolla_config: openstack_release: ocata customize: nova.conf: | [oslo_messaging_notifications] driver = messagingv2 [DEFAULT] disk_allocation_ratio = 20.0 ram_allocation_ratio = 5.0 cpu_allocation_ratio = 25.0 1. This VNC params in nova.conf is not taking effects. This makes it not possible to directly access instance’s VNC console, from both Horizon/ContrailCommand. [vnc] enabled = True novncproxy_base_url = http://127.0.0.1:20002/vnc_auto.html In details: * For the current manual setup with 5.0.1, if I include the above params, I can directly access instance console from my web browser, by port map 192.168.250.1:6080 (remote VNC port) to 127.0.0.1:20002 (local) (Note the URL for console is 127.0.0.1:20002, straight from Horizon) [cid:image001.png@01D445C5.5F25BE90] But if I use Contrail Command, with the same params in nova.conf, it not taking effect, and console_vnc_url still pointing to 192.168.250.1:6080 and timeout. This make direct console access not possible (from both OpenStack and Contrail Command) [cid:image002.png@01D445C6.815CC9D0] If I manually change the above URL to 127.0.0.1:20002, console works. So this novncproxy_base_url param under /etc/kolla/*/nova.conf is not in effect. I reboot the hosts several time already. [cid:image003.png@01D445C6.BCF2D8E0] How should I resolve those issues? Thank you. Regards, Linh From: Ali Zaringhalam Sent: Tuesday, September 4, 2018 3:13 PM To: Ashutosh Gupta <>>; Chaitanya Kadiyala <>>; Linh Phan <>> Cc: Dilip Sundarraj <>> Subject: RE: 5.0.01 /Try Upgrade and Documentation - Working Session Hi Ashutosh: Thanks for your investigation. We will increase memory to 64 GB and test again to see if we see the issue. However I would like to have some clarification regarding memory as the root cause. In 3.2.3 we had 24 G of RAM and it served us quite well with all aspects of the setup including Cassandra. While 5.0.1 as a whole may require more memory (we increased it to 32 G and still have some headroom in “top”), it is not clear why Cassandra by itself requires so much more RAM in 5.0.1, and then only intermittently? In 3.2.3 we ran 5 PODs on a single server and are able to support 30 PODs on 6 servers. In 5.0.1 we reduced the count to 4 (and a total of 6x4=24 PODs) because of higher memory requirements and the addition of a second compute node. If we are required to assign 64 G to the config node alone, we will have to reduce the count still further to 3 PODs per server. With our 6 servers currently in place, the number of PODs will be reduced from the current 30 in 3.2.3 to 18 in 5.0.1. Given the significant resource impact I wonder if there are parameters that can be tweaked to manage this better for a non-production environment such as ours? Regards, Ali From: Ashutosh Gupta Sent: Tuesday, September 4, 2018 4:27 PM To: Chaitanya Kadiyala <>>; Linh Phan <>> Cc: Ali Zaringhalam <>>; Dilip Sundarraj <>> Subject: Re: 5.0.01 /Try Upgrade and Documentation - Working Session Hello Chaitanya The Cassandra state down issue is resolved and the problem was low memory on the config node. It is advisable to increase the memory from 32 Gb to 64 Gb on the config node to avoid this issue in the future. Also as Sarin suggested you can fresh install Contrail Command 5.0.1. Thanks, Ashutosh ________________________________ From: Chaitanya Kadiyala Sent: Friday, August 31, 2018 12:35:04 PM To: Dilip Sundarraj; Ashutosh Gupta; Arohi Gupta; Linh Phan Cc: Ali Zaringhalam Subject: 5.0.01 /Try Upgrade and Documentation - Working Session When: Tuesday, September 4, 2018 11:00 AM-12:00 PM. Where: Skype Meeting Hi Ali, Dilip and his team are on-board in helping with expediting the documentation for 5.0.1 tasks for the sandbox as well as help with debugging the upgrade including COntrial Command. I want to bring his team and yours together in this call to address any potential gaps in the existing upgrade. Hope the timing works for you all. Thanks Chaitanya ......................................................................................................................................... Join Skype Meeting Trouble Joining? Try Skype Web App Join by phone +14089369000 (USA, West Coast) English (United States) +19785898300 (USA, West Coast) English (United States) ILYNC (45962) (USA, West Coast) English (United States) +18446454398 (USA, West Coast) English (United States) +18332135669 (USA, West Coast) English (United States) Find a local number Conference ID: 58410349 Forgot your dial-in PIN? |Help Please consider with whom you are communicating, including non-Juniper personnel of companies federated on Skype for Business, before sharing any confidential information. Non-disclosure agreements may apply. [!OC([1033])!] ......................................................................................................................................... _____________________________________________ From: Chaitanya Kadiyala Sent: Friday, August 31, 2018 12:26 PM To: Dilip Sundarraj <>>; Ashutosh Gupta <>>; Arohi Gupta <>> Cc: Yusuke Kanamori <>>; Ali Zaringhalam <>>; Jerry Passione <>> Subject: FW: 5.0.01 /Try pod 26 to 30 SSH keys Hi Dilip, Ashutosh, Arohi, We have 5.0.1 installed with the following infrastructure. • 2 compute nodes to support VM migration. • JTI telemetry from the vMX gateway to AppFormix collector. Traffic from the overlay VMs (outbound towards internet; to/from linux jump server with floating IP) can be viewed in AppFormix dashboard for vMX. • ssh port-forwarding through the Linux jump server. GUI access via local browser. (RDP will no longer be supported.) • As before, full support for floating IP with outbound internet access from overlay. ssh access to overlay from the Linux jump server. • All legacy images are included. In addition, a vSRX 15.1X49-D140.2 (in factory default) and Kali Linux images are available. You can logon to Pod 28 (access keys listed in the attached email) to see the current build and stability of the setup. We do need help from your team in expediting the Contrail Command install as well as in updating the documentation for the sandbox. I’ve cc’ed Ali from OpenLab who is driving the upgrade and facilitating the infra for the sandbox. I will setup a call between all of us for Tuesday to go over the details for this sandbox upgrade. Thanks Chaitanya From: Laurent Paumelle Sent: Thursday, August 23, 2018 11:33 AM To: Ali Zaringhalam <>>; Stuart Mackie <>>; Linh Phan <>>; Fredric Linder <>>; Chaitanya Kadiyala <>>; Chameera Wijesinghe <>> Cc: Scott Alexander <>>; Jerry Passione <>>; Marcelino Chua <>>; Maria Caraballo <>> Subject: Re: 5.0.01 /Try pod 26 to 30 SSH keys That’s fine for me ☺ it was just a generic question as other people may also need to put their own software sometimes on a pod. Thanks for the setup! Laurent From: Ali Zaringhalam <>> Date: Thursday 23 August 2018 at 10:14 To: Laurent Paumelle <>>, Stuart Mackie <>>, Linh Phan <>>, Fredric Linder <>>, Chaitanya Kadiyala <>>, Chameera Wijesinghe <>> Cc: Scott Alexander <>>, Jerry Passione <>>, Marcelino Chua <>>, Maria Caraballo <>> Subject: RE: 5.0.01 /Try pod 26 to 30 SSH keys We should consider crx when we build k8/contrail. otherwise it is obsolete. Regards, Ali From: Laurent Paumelle Sent: Thursday, August 23, 2018 12:45 PM To: Ali Zaringhalam <>>; Stuart Mackie <>>; Linh Phan <>>; Fredric Linder <>>; Chaitanya Kadiyala <>>; Chameera Wijesinghe <>> Cc: Scott Alexander <>>; Jerry Passione <>>; Marcelino Chua <>>; Maria Caraballo <>> Subject: Re: 5.0.01 /Try pod 26 to 30 SSH keys Actually there WAS a plugin at the Contrail 3.x timeframe named nova-docker for Openstack. With this any container could be launched directly like a regular VM in the UI/Heat template. Later with Contrail 4.x I used instead a script to integrate and launch cSRX and create the interface mapping to the vRouter. I do not remember yet but I think the Nova-Docker was already deprecated and no replacement ready. Anyway that script was made by a Contrail developer and helped a lot. Now with Contrail 5.x and newer Openstack, I’ll need to look whether a plugin exists to integrate with containers but I believe that yes we may need to install Contrail/K8 in that case. cSRX does not work with K8 per say as it would then need “CNN” (multiple interface support as opposition to CNI for single interface) but it does work thanks to Contrail vRouter since it’s mapped to it rather than to the native K8 networking and Contrail does the service chaining between vnetworks with cSRX or any VNF of course. It does solve this puzzle ☺ Laurent From: Ali Zaringhalam <>> Date: Thursday 23 August 2018 at 09:20 To: Laurent Paumelle <>>, Stuart Mackie <>>, Linh Phan <>>, Fredric Linder <>>, Chaitanya Kadiyala <>>, Chameera Wijesinghe <>> Cc: Scott Alexander <>>, Jerry Passione <>>, Marcelino Chua <>>, Maria Caraballo <>> Subject: RE: 5.0.01 /Try pod 26 to 30 SSH keys Not sure if there is “and may be some docker plugin for Openstack/Contrail”. cSRX should probably be done with K8/Contrail. Regards, Ali From: Laurent Paumelle Sent: Thursday, August 23, 2018 12:17 PM To: Ali Zaringhalam <>>; Stuart Mackie <>>; Linh Phan <>>; Fredric Linder <>>; Chaitanya Kadiyala <>>; Chameera Wijesinghe <>> Cc: Scott Alexander <>>; Jerry Passione <>>; Marcelino Chua <>>; Maria Caraballo <>> Subject: Re: 5.0.01 /Try pod 26 to 30 SSH keys Thanks Ali and Linh for the setup. Quick question: how complex is it to customize a Pod with additional software? For example, I’m also planning to test with cSRX at some point and then will need Docker (and may be some docker plugin for Openstack/Contrail) and then I may need to try adding some additional components, but I do not want to break things of course. Then my question on how complex it is and if possible at all? Thanks Laurent From: Ali Zaringhalam <>> Date: Thursday 23 August 2018 at 08:17 To: Stuart Mackie <>>, Linh Phan <>>, Fredric Linder <>>, Laurent Paumelle <>>, Chaitanya Kadiyala <>>, Chameera Wijesinghe <>> Cc: Scott Alexander <>>, Jerry Passione <>>, Marcelino Chua <>>, Maria Caraballo <>> Subject: RE: 5.0.01 /Try pod 26 to 30 SSH keys • Based on feedback from folks on this thread we need to install Contrail Command. • contrail-command and contrail-command-deployer are available from the GA registry but with a different tag (simple 5.0.1-0.214 instead of 5.0.1-0.214-ocata). We don’t need to go to a separate registry. It needs a separate installation. • We will install Contrail Command on the Linux jump server. This obviates the need for a separate VM to host Contrail Command. It will be port-mapped. Regards, Ali From: Stuart Mackie Sent: Thursday, August 23, 2018 1:44 AM To: Linh Phan <>>; Ali Zaringhalam <>>; Fredric Linder <>>; Laurent Paumelle <>>; Chaitanya Kadiyala <>>; Chameera Wijesinghe <>> Cc: Scott Alexander <>>; Jerry Passione <>>; Marcelino Chua <>>; Maria Caraballo <>> Subject: Re: 5.0.01 /Try pod 26 to 30 SSH keys Definitely, it will be the default GUI for most configuration actions. Stuart -914 886 2534 From: Linh Phan <>> Date: Thursday, August 23, 2018 at 7:20 AM To: Stuart Mackie <>>, Ali Zaringhalam <>>, Fredric Linder <>>, Laurent Paumelle <>>, Chaitanya Kadiyala <>>, Chameera Wijesinghe <>> Cc: Scott Alexander <>>, Jerry Passione <>>, Marcelino Chua <>>, Maria Caraballo <>> Subject: RE: 5.0.01 /Try pod 26 to 30 SSH keys Hi Stuart, We use the standard Contrail-ansible-deployer to deploy this setup: https://github.com/Juniper/contrail-ansible-deployer Contrail-Command-Deployer is a separate setup and is not included with the standard deployer. I checked port 9091 (no response) and no Contrail-Command container is present. https://github.com/Juniper/contrail-command-deployer We will need to experiment and test Contrail-Command on a separate setup. Do you suggest we put Contrail-Command for the /try as well? Regards, Linh. From: Stuart Mackie Sent: Wednesday, August 22, 2018 10:02 PM To: Ali Zaringhalam <>>; Fredric Linder <>>; Laurent Paumelle <>>; Chaitanya Kadiyala <>>; Chameera Wijesinghe <>> Cc: Scott Alexander <>>; Jerry Passione <>>; Marcelino Chua <>>; Maria Caraballo <>>; Linh Phan <>> Subject: Re: 5.0.01 /Try pod 26 to 30 SSH keys Is there a port map for Contrail Command? Stuart -914 886 2534 From: Ali Zaringhalam <>> Date: Thursday, August 23, 2018 at 4:13 AM To: Stuart Mackie <>>, Fredric Linder <>>, Laurent Paumelle <>>, Chaitanya Kadiyala <>>, Chameera Wijesinghe <>> Cc: Scott Alexander <>>, Jerry Passione <>>, Marcelino Chua <>>, Maria Caraballo <>>, Linh Phan <>> Subject: 5.0.01 /Try pod 26 to 30 SSH keys Hi Folks: The 5.0.1 release is now available for testing. You can use the same ssh keys we sent earlier for the 5.0 setups. 1. POD 27 (DC-198): Stuart Mackie 2. POD 28 (DC-204): Chaitanya Kadiyala 3. POD 29 (DC-210): Fred Linder/Chameera Wijesinghe @Laurent: we are setting you up on another server (POD25) and will send you the coordinates. Please note that due to the increase in the number of compute nodes, we can only fit 4 PODs per ESXi server. Given the 6 servers at our disposal, we can only support 24 PODs instead of the previous 30. We are working to increase the number of servers. Regards, Ali _____________________________________________ From: Ali Zaringhalam Sent: Monday, July 9, 2018 11:13 PM To: Stuart Mackie <>>; Fredric Linder <>>; Laurent Paumelle <>>; Chaitanya Kadiyala <>>; Chameera Wijesinghe <>> Cc: Scott Alexander <>>; Jerry Passione (>) <>>; Marcelino Chua <>>; Maria Caraballo <>>; Linh Phan <>> Subject: FW: /Try pod 26 to 30 SSH keys Hi Folks: The Contrail 5.0 sandboxes are now available for testing in the JCL /try. The primary purpose of making these available early is to test existing use cases and user experience. The target general availability in all 30 sandboxes is 8/15. This is contingent on incorporating your early feedback and the availability of user documentation for both Contrail and AppFormix. The sandboxes are allocated as follow: 1. POD 26 (DC-192): Linh Phan 2. POD 27 (DC-198): Stuart Mackie 3. POD 28 (DC-204): Chaitanya Kadiyala 4. POD 29 (DC-210): Fred Linder/Chameera Wijesinghe 5. POD 30 (DC-216): Laurent Paumelle The following new capabilities are available: • 2 compute nodes to support VM migration. • JTI telemetry from the vMX gateway to AppFormix collector. Traffic from the overlay VMs (outbound towards internet; to/from linux jump server with floating IP) can be viewed in AppFormix dashboard for vMX. • ssh port-forwarding through the Linux jump server. GUI access via local browser. (RDP will no longer be supported.) • As before, full support for floating IP with outbound internet access from overlay. ssh access to overlay from the Linux jump server. • All legacy images are included. In addition, a vSRX 15.1X49-D140.2 (in factory default) and Kali Linux images are available. The intent of the sandbox is to provide a platform to support OpenStack/Contrail/AppFormix features. The primary target users are external reservations and Contrail/AppFormix training classes. Users/trainers are responsible for the development and testing of their use cases. We plan to provide similar sandboxes with expanded capabilities and privileges for internal COE/SE users in JCL BYO in September. Until that time, the use of these sandboxes for external demos should be considered as temporary. Let us know if you have any questions. Regards, Ali _____________________________________________ From: Linh Phan Sent: Monday, July 9, 2018 6:34 PM To: Ali Zaringhalam <