- Announce - mailman3.cels.anl.gov

Re: [Announce] New Chameleon hardware available
by Pierre Riteau 14 Mar '18

14 Mar '18

We are also happy to announce that stitching is now available on CHI@TACC! For more details see our documentation: https://www.chameleoncloud.org/docs/bare-metal-user-guide/stitching-chamele… > On 14 Mar 2018, at 16:46, Pierre Riteau <priteau(a)uchicago.edu> wrote: > > We are pleased to announce immediate availability of the first installment of new hardware for phase 2 of Chameleon! > > This new hardware is composed of two 32-node racks of Skylake nodes. Each node is a PowerEdge R740 server with two Intel Xeon Skylake CPUs running at 2.60 GHz (each with 12 cores / 24 threads), 192 GiB of RAM, a Samsung SSD with 240 GB of storage, and 10 Gigabit Ethernet connectivity. 32 of those bare metal nodes are available on CHI@UC and 32 are on CHI@TACC. > > The nodes are connected at 10 Gigabit to Corsa switches: on CHI@TACC, to a Corsa DP2200, and on CHI@UC, to a Corsa DP2400. Both are using a 100 Gigabit uplink; at TACC, it is connected to a 100G network while at UC, connectivity to a 100G network will soon be available. Corsa switches offer advanced virtualization and OpenFlow capabilities which we plan to leverage in the near future to support user controlled networking experiments—for now, these switches are conventionally configured to support multi-tenant networking. Please note, however, that on CHI@UC the only isolated networks the new nodes can use are the VLANs stitchable to ExoGENI. We are working on allowing use of the full range of VLANs in the future. > > The new nodes can be reserved using the "compute_skylake" node type on the command line or by selecting "Compute (Skylake)" in the web interface. For consistency, the existing compute nodes have been renamed to "compute_haswell" and Infiniband nodes to "compute_haswell_ib": please adapt your scripts accordingly. > > We look forward to learn about all the exciting new experiments enabled by this new hardware—happy experimenting!

1 0

New Chameleon hardware available
by Pierre Riteau 14 Mar '18

14 Mar '18

We are pleased to announce immediate availability of the first installment of new hardware for phase 2 of Chameleon! This new hardware is composed of two 32-node racks of Skylake nodes. Each node is a PowerEdge R740 server with two Intel Xeon Skylake CPUs running at 2.60 GHz (each with 12 cores / 24 threads), 192 GiB of RAM, a Samsung SSD with 240 GB of storage, and 10 Gigabit Ethernet connectivity. 32 of those bare metal nodes are available on CHI@UC and 32 are on CHI@TACC. The nodes are connected at 10 Gigabit to Corsa switches: on CHI@TACC, to a Corsa DP2200, and on CHI@UC, to a Corsa DP2400. Both are using a 100 Gigabit uplink; at TACC, it is connected to a 100G network while at UC, connectivity to a 100G network will soon be available. Corsa switches offer advanced virtualization and OpenFlow capabilities which we plan to leverage in the near future to support user controlled networking experiments—for now, these switches are conventionally configured to support multi-tenant networking. Please note, however, that on CHI@UC the only isolated networks the new nodes can use are the VLANs stitchable to ExoGENI. We are working on allowing use of the full range of VLANs in the future. The new nodes can be reserved using the "compute_skylake" node type on the command line or by selecting "Compute (Skylake)" in the web interface. For consistency, the existing compute nodes have been renamed to "compute_haswell" and Infiniband nodes to "compute_haswell_ib": please adapt your scripts accordingly. We look forward to learn about all the exciting new experiments enabled by this new hardware—happy experimenting!

1 0

Merry Christmas from Chameleon!
by Kate Keahey 23 Dec '16

23 Dec '16

Merry Christmas from Chameleon! … and we couldn’t finish the year without putting a little something under the tree, this time on a power management theme — we are trying save all the lumps of coal we can for the stockings! Down the chimney come the following: New low power hardware. We made available 16 new servers from the HPE Moonshot family: 8 low power Xeon servers (HP ProLiant m710p with one 4-core Intel Xeon E3-1284L v4 processor) and 8 Atom servers (HP ProLiant m300 with one 8-core Intel Avoton-based System on a Chip). This new low power hardware is particularly well suited for experiments on energy-efficient computing. The new nodes are available on CHI@TACC ; the Atom nodes are compatible with our supported CentOS 7 and Ubuntu images however the Xeon nodes do not yet support CentOS (we are working on it with the manufacturer and hope to work this out soon). The nodes can be discovered via our resource discovery interface <https://www.chameleoncloud.org/user/discovery/> and can be selected when creating reservations <https://chi.tacc.chameleoncloud.org/dashboard/project/leases/>. Power management tools. To facilitate experiments with power management we are making available a power measurement tool <https://github.com/coolr-hpc/intercoolr> as part of our base images. This tool leverages APIs from Intel processors to measure power consumption of each CPU socket and of memory DIMMs. We would like to thank Chameleon user Kazutomo Yoshii <http://www.mcs.anl.gov/person/kazutomo-yoshii> of Argonne National Laboratory <http://www.anl.gov/> for developing and sharing this tool. To learn how to use it, read our documentation at https://www.chameleoncloud.org/monitor-and-collect <https://www.chameleoncloud.org/monitor-and-collect>. This utility is compatible with all our hardware except for the Intel Atom nodes, but we are hoping to extend support for them in the future. Enjoy the new toys -- and most of all have a safe and happy Holiday Season and a fantastic New Year! Kate Keahey Mathematics and CS Division, Argonne National Laboratory Computation Institute, University of Chicago

1 0

Happy Halloween from Chameleon!
by Kate Keahey 31 Oct '16

31 Oct '16

In the true Halloween spirit we declared war on – and vanquished (or at least did severe damage to) – several dangerous demons that we are all afflicted by on a daily basis: The Ghost of Complexity Past. Many of our users want to deploy “virtual clusters”, such as for example OpenStack installations or MPI clusters. They are often hard to deploy because their configuration involves exchange of information typically available only at deployment time, such as hostnames or security keys. These last configuration steps are therefore often carried out manually, increasing complexity and making reproducibility hard. To solve this problem we have deployed a tool for configuring such complex appliances so that they can be deployed automatically. Our complex appliance deployment orchestrator allows you to configure complex “virtual clusters” on bare metal hardware “with one click”. It takes an image (appliance) and a template – a document that defines how many instances of this appliance to deploy, what information to exchange, and what scripts to run on deployment – and then deploys a cluster of specified size automatically, and in exactly the same way, time and time again. It also supports parameterizing deployments – for example, configuring the number of nodes of a cluster. Complex appliance support is currently available on CHI@UC and CHI@TACC. To get started, please check out our guide <https://www.chameleoncloud.org/docs/complex-appliances/> to working with complex appliances. The time-sucking Vampire of Having-to-Do-Everything-Yourself. While we are confident that many of you will want to configure your own complex appliances representing interesting frameworks you have developed, we thought we’d get you started with a few. They include several of the most popular distributed applications and frameworks: OpenStack Mitaka <https://www.chameleoncloud.org/appliances/30/>: this complex appliance deploys OpenStack Mitaka with DevStack over one controller node and a configurable number of compute nodes MPI bare-metal cluster <https://www.chameleoncloud.org/appliances/29/>: this complex appliance deploys a bare-metal MPI cluster using the MVAPICH2 implementation NFS share <https://www.chameleoncloud.org/appliances/25/>: this complex appliance deploys an NFS server exporting a directory to a configurable number of NFS clients MPI + SR-IOV KVM cluster <https://www.chameleoncloud.org/appliances/28/>: this complex appliance deploys an MPI cluster of KVM virtual machines using the MVAPICH2-Virt implementation and configured with SR-IOV for high-performance communication over Infiniband All of these complex appliances can be easily found in our Appliance Catalog <http://www.chameleoncloud.org/appliances>; they are identified by a cluster icon in the top right corner. We hope that you will configure and share with other users complex appliances that represent the research you are working on using our instructions for customizing or writing new complex appliances <https://www.chameleoncloud.org/docs/complex-appliances/#toc-customizing-an-…> and sharing them <https://www.chameleoncloud.org/docs/complex-appliances/#toc-sharing-new-com…>. The Nightmare of Knowing-about-Cool-New-Hardware-without-Being-Able-to-Use-It. Now you can lay your hands on non-volatile memory and FPGAs! We added two NVMe SSDs to each of our two storage hierarchy nodes. Each device is a 2.0 TB Intel SSD DC P3700, advertised with a sequential read throughput of 2,800 MB/s and a sequential write throughput of 2,000 MB/s. The storage hierarchy available on each of the storage nodes is now: 512 GiB of RAM, two NVMe SSDs of 2.0 TB each, four Intel S3610 SSDs of 1.6 TB each, and four 15K SAS HDDs of 600 GB each. We also made available four FPGA nodes. Each of these nodes is fitted with a Nallatech 385A board <http://www.nallatech.com/store/pcie-accelerator-cards/nallatech-385a-arria1…> with an Altera Arria 10 1150 GX FPGA (up to 1.5 TFlops), 8 GiB DDR3 of on-card memory, and dual QSFP 10/40 GbE support. They are configured to run OpenCL code, but they can be reconfigured (by a request to our help desk) to run compiled designs prepared with Altera Quartus. Due to export control limitations, access to the development toolchain requires verification of your user profile. We explain how to gain access to the development toolchain and execute code on the FPGA nodes in the FPGA guide <https://www.chameleoncloud.org/docs/bare-metal-user-guide/fpga/>. The Specter of Too-High-Entry-Barrier. We replaced the dusty old spellbook that served us all so well with a fresh bare metal guide that is much easier to follow for different groups of users. We have now split the documentation <https://www.chameleoncloud.org/docs/bare-metal-user-guide/> into two parts, one focusing on using Chameleon through its web interface, and the other focusing on using Chameleon through command-line and APIs. We also created a Community Resources <https://www.chameleoncloud.org/docs/community/> space on the Chameleon website (currently available off the Documentation tab) and added two new training videos demonstrating how to use Chameleon for experimental research with a use case showing a performance evaluation of Spark – one for Linux and OS X users <https://www.youtube.com/watch?v=J6UPmT41gq8> and one for Windows users <https://www.youtube.com/watch?v=FRGehi9UVG4>. From now on, this will be the space where education and training materials using Chameleon will get released – we know that many of you are using Chameleon in classes and hope that these materials will be useful. As always, we are interested in your feedback, so please let us know <https://www.chameleoncloud.org/user/help/ticket/new/> what types of training materials you would like to see or if you have any you would like to contribute! Happy Halloween! Kate Keahey Mathematics and CS Division, Argonne National Laboratory Computation Institute, University of Chicago

1 0

Happy Independence Day from Chameleon!
by Kate Keahey 01 Jul '16

01 Jul '16

When in the Course of human events, one people take upon themselves to nourish and support the research needs of another, a decent respect for opinions of mankind requires that, in well-appointed intervals, a declaration of the works performed should be made to those whose research can profit by them. We hold these truths to be self-evident: More types of hardware will support a broader range of experiments. We have made available new hardware consisting of two storage hierarchy nodes, two K80 GPU nodes, and two M40 GPU nodes. Each of the additional six nodes is a Dell PowerEdge R730 server with the same CPUs as our compute nodes. The two storage hierarchy nodes have been designed with enabling experiments using multiple layers of caching in mind: they are configured with 512 GB of memory, four Intel S3610 SSDs of 1.6 TB each, and four 15K SAS HDDs of 600 GB each. The GPU nodes are targeting experiments using accelerators to improve the performance of some algorithms, experiments with new visualization systems, and deep machine learning. Each K80 GPU node is upgraded with an NVIDIA Tesla K80 accelerator, consisting of two GK210 chips with 2496 cores each (4992 cores in total) and 24 GB of GDDR5 memory. Each M40 node is upgraded with an NVIDIA Tesla M40 accelerator, consisting of a GM200 chip with 3072 cores and 24 GiB of GDDR5 memory. In order to make it easy for users to get started with the GPU nodes, we have developed a CUDA appliance <https://www.chameleoncloud.org/appliances/21/> that includes NVIDIA drivers as well as the CUDA framework. For more information on how you can reserve these nodes, see the heterogeneous hardware section <https://www.chameleoncloud.org/docs/bare-metal-user-guide/#heterogeneous_ha…> of the bare metal user’s guide. For a limited time, we have an “opening sale” on these nodes: right now they will cost you the same in SUs as regular service nodes while offering significantly more powerful capabilities. Last but not least, we would like to thank Early User Luc Renambot <https://www.evl.uic.edu/> from University of Illinois in Chicago for early testing and providing useful feedback on our new hardware. Big Data experiments require a large data store close to the testbed. To facilitate this we have developed a new Chameleon object store service. The object store can be accessed via the OpenStack Swift interface and is currently backed by a 1.6 PB Ceph cluster with double replication, effectively providing 0.8 PB for general use. We intend to enlarge this storage capacity as needed, but for now, here too we have an “opening sale”: for a limited time there are no limits on usage. The object store is located at TACC, but available to users of both CHI and KVM resources at TACC, UC, and anywhere else on the Internet. To make it easier to use, we added a Swift client to all appliances supported by Chameleon. More details can be found in the object store section of the bare-metal documentation guide <https://www.chameleoncloud.org/docs/bare-metal-user-guide/#object_store>. Better appliance management tools make experimentation easier and more fun. We have put in place a process that allows us to easily create new images as well as maintain, and customize them -- and we are now making it available to you. Our process uses an adaptation of the OpenStack diskimage-builder <https://github.com/openstack/diskimage-builder> to make it compatible with the Chameleon testbed. We use this process to create, support, and customize appliances available via the appliance catalog <https://www.chameleoncloud.org/appliances/>. We have made the scripts to generate the base appliances (CC-CentOS7, CC-Ubuntu14.04, CC-Ubuntu16.04) available, so you can easily fork them to create your own customizations of base images. For more information see our documentation on Chameleon Image Builder <https://www.chameleoncloud.org/docs/bare-metal-user-guide/#toc-building-and…>. Further, having this process in place has allowed us to develop an image that works for both TACC and UC clouds (whole disk image format only). This means that you can snapshot your instance at TACC, transfer it to UC, and launch it there without making any changes -- which can be very handy if you need to move your work from site to site quickly. And talking of snapshotting, we improved that too: we now have a tool, called cc-snapshot <https://github.com/ChameleonCloud/ChameleonSnapshotting>, that allows you to save the instance you are working on, with all its modifications, and upload a copy to Glance; what used to take multiple complicated commands can now be done quickly with just one command. Having more ready to use appliances allows scientists to spend more time on research. We immediately put the Chameleon Image Builder to good use and created (and even already updated) multiple new appliances that can be found in the appliance catalog <https://www.chameleoncloud.org/appliances/>. In addition to the CUDA appliance <https://www.chameleoncloud.org/appliances/21/> described above, we also added the Ubuntu appliance <https://www.chameleoncloud.org/appliances/19/> (and already upgraded it to 16.04), upgraded the CentOS 7 appliance <https://www.chameleoncloud.org/appliances/1/>, added the SR-IOV MVAPICH2-Virt <https://www.chameleoncloud.org/appliances/9/> appliance that packages efficient MPI for KVM running over IB, SR-IOV RDMA-Hadoop <https://www.chameleoncloud.org/appliances/17/> appliance that packages efficient Hadoop over IB, and last but not least the Ubuntu 14.04 DevStack Mitaka <https://www.chameleoncloud.org/appliances/11/> which packages OpenStack. Making it easy to learn about Chameleon will make experimental capabilities more accessible. In order to make it easier for users to use Chameleon we have developed and scheduled a series of webinars <https://www.chameleoncloud.org/user/webinar> that provide a hands-on, in-depth introduction to running your experiments on bare metal in only an hour. Each webinar will guide users through completing a small Computer Science research project using Chameleon and will provide support every step of the way. It is designed to introduce a workflow that can be easily extended to support the user’s own research experimentation. The current schedule will be extended on demand. Have a fantastic Independence Day and may your pursuit of happiness be fruitful! Kate Keahey Mathematics and CS Division, Argonne National Laboratory Computation Institute, University of Chicago

1 0

From Chameleon, with love
by Kate Keahey 12 Feb '16

12 Feb '16

We love our users! That’s why for the last few months we have been working on a Valentine. Not the pretty picture – rather, we’d like to lay at your feet a bouquet of new capabilities that we hope will make your research easier and more productive. These include: - Support for custom kernel boot. You can now easily customize the operating system kernel or modify the kernel command line in your experiments! While it was possible (just barely) to boot from a custom kernel before, it was also inefficient and slow for those of you who needed to do it repeatedly: each time, a user had to upload a new image with the customized kernel to the repository and then boot from that image; both time-consuming operations. You now have the option of modifying the boot loader configuration to point it to a new kernel on the local disk, or specifying kernel parameters and then rebooting using this modified configuration. This will make updates to the kernel take effect quickly and means that experiments requiring kernel development will be much easier to run. If you are not working with kernels, you will see no changes in how images are deployed – however, we have had to change image format of bare-metal images (from partition images to whole disk images) as well as image snapshotting instructions <https://www.chameleoncloud.org/docs/user-guides/ironic/#snapshotting_an_ins…>. To minimize the impact on you, we have converted all of the images supported by Chameleon and as many of the user images as we could; in order to convert others we may need your assistance on Monday. Note that your existing images have been renamed with a “.partition” suffix. If you would like to find out more about the feature, please read our bare metal user guide <https://www.chameleoncloud.org/docs/user-guides/bare-metal-user-guide/#toc-…>. We would like to thank our users Swann Perarnau <http://www.mcs.anl.gov/~perarnau/> of Argonne National Laboratory and Yuyu Zhou <http://people.cs.pitt.edu/~yuyuzhou/> of the University of Pittsburgh for helping us define the details of this capability and working with us to test and validate them! - Upgrade to Liberty. The OpenStack deployment which is a core component of CHI (Chameleon Infrastructure) has been upgraded to the most recent Liberty release. This upgrade not only facilitated the development of the feature above, but also added multiple usability features, such as the ability to edit compute image metadata from the project dashboard or the ability to deactivate an image, improved error messaging, improved performance of many features, and added administrative features that will allow us to make Chameleon operate more smoothly. - Appliance Marketplace. The appliance marketplace allows you to discover, publish, and share appliances – bare metal or virtual machine images capturing the experimental environment – a key element of reproducibility. The initial version of the Appliance Marketplace announced in our New Year’s message was simply a static table of initial appliances provided and supported by the Chameleon team. The new Appliance Marketplace allows you to publish and share your own appliances so that others can discover them and easily play with the software and tools you developed. They can then use it either to reproduce the research of others or as a starting point for their own research and experimentation. To find out more, read the appliance documentation on our FAQ <https://www.chameleoncloud.org/docs/user-faq/#toc-appliances>. We would like to thank Rosa M. Badia and her team <http://www.bsc.es/computer-sciences/grid-computing> of the Barcelona Supercomputing Center for motivating us to provide this capability sooner rather than later and for contributing the first external appliance representing COMPSs <http://compss.bsc.es/>, a task based programming model for distributed platforms! - Identity-based federation with GENI <http://www.geni.net/>. For some time now, GENI users have been able to access Chameleon by using their GENI credentials and charging to the GENI Federation Project. Now Chameleon users will be able to use their Chameleon credentials to access GENI without having to create a GENI account! This will enable multiple additional advanced research projects. We would like to thank Tom Mitchell <http://www.geni.net/?page_id=1927> and Marshall Brinn <http://www.geni.net/?page_id=2080> of BBN on the GENI team for working with us to make this possible! As most of you already know, these upgrades have already been applied at the University of Chicago site, and will be applied at TACC at the scheduled downtime next week. We hope that these new features will reduce the time you spend wrangling logistics and put it back where it belongs: into your research. Thank you to all for sharing your requirements with us and inspiring us to make Chameleon a better testbed! If you find Chameleon useful for your research, send us a Valentine at users(a)chameleoncloud.org <mailto:[email protected]>! Here’s what we would like: a story of how you are using Chameleon for your research. Those stories are motivating not only to us in that they make us build a better testbed – but also to others who may get good ideas from your work on how to structure their own experiments. To encourage this sharing of experiences, we plan to highlight interesting research projects done by our users on the Chameleon portal as we have recently done with the story on cybersecurity research <https://www.chameleoncloud.org/news/research-highlight-circumventing-cyber-…>. In short, we’d like to challenge you to be an inspiration to other: if you have interesting stories to share, email them to users(a)chameleoncloud.org <mailto:[email protected]>! Happy Valentine’s Day! Kate Keahey Mathematics and CS Division, Argonne National Laboratory Computation Institute, University of Chicago

1 0

Happy New Year from Chameleon!
by Kate Keahey 06 Jan '16

06 Jan '16

Happy New Year from Chameleon! We’ve had a fantastic 2015 building the testbed -- many thanks to the Chameleon community for your patience and support! The last year focused on making a large-scale, highly reconfigurable testbed for Computer Science possible. A few highlights: - The release of Technology Preview of our bare metal reconfiguration capabilities (CHI) in early April, - Public release of CHI on new hardware in late July, - Making a new OpenStack KVM cloud available in September. As of today Chameleon has 500+ users working on almost 150 exciting projects ranging from low-level operating systems research to innovative projects in cloud computing education. The more we learn about your individual activities the more impressed we are with the depth and breadth of all the exciting projects going on! One major addition in last quarter of 2015 was the introduction of identity federation with GENI. GENI users can now use their GENI credentials to log into the Chameleon portal. This allows them to access Chameleon without having to register a separate account, via a a shared GENI federation project with enough SUs to try out Chameleon. GENI users can of course also create individual Chameleon projects dedicated to their research. In 2016 our goal is to take Computer Science research on Chameleon from possible to easy. Here is a preview of some of the exciting features we are planning for the first months of 2016: - Chameleon appliances. We will improve ways of creating and sharing Chameleon appliances, i.e., the bare metal images that you deploy to create your environment. We will provide better ways of snapshotting appliances, developing complex appliances, and build an appliance marketplace where users will be able to publish their own appliances and discover appliances contributed by others. In fact, to celebrate New Year’s, we just released a small preview of the marketplace <https://www.chameleoncloud.org/docs/appliances/> with some initial appliances. - Experiment management. We will provide tools that will allow you to automate your experiments so that they are easy to deploy and manage even when you are not around. This will also make it easy to repeat them as many times as needed – or to pass them on to somebody else in your group or outside of it to reproduce. - Deeper reconfiguration and better monitoring. We will provide deeper reconfiguration and monitoring capabilities making it easier for users working on operating systems research or otherwise requiring deep reconfigurability. Many more ideas are coming down the pipe and we will keep you posted – in the meantime, please share your suggestions with us and tell us how we can help you use Chameleon better. All in all, we are looking forward to another exciting year! Kate Keahey Mathematics and CS Division, Argonne National Laboratory Computation Institute, University of Chicago

1 0

New OpenStack KVM cloud available
by Pierre Riteau 28 Sep '15

28 Sep '15

We are happy to announce that the new Chameleon OpenStack KVM cloud is available! This cloud spans three additional racks of Chameleon hardware with configuration similar to our bare metal racks (see https://www.chameleoncloud.org/about/hardware-description/ for more details about our hardware).The cloud is configured with the most recent OpenStack release (Kilo) and offers the same set of OpenStack services as the KVM cloud available up to now. All images, snapshots, volumes, and volumes snapshots have been copied today from the old hardware to the new cloud. However, no instances have been migrated. If you are already a Chameleon user, follow the OpenStack KVM User Guide (https://www.chameleoncloud.org/docs/user-guides/openstack-kvm-user-guide/) to access the new cloud; if you are new to Chameleon get started in three simple steps at https://www.chameleoncloud.org/docs/getting-started/ If required, you can continue using the old hardware, but please note that it will become unavailable after October 9. We also won't be able to migrate user data (images, snapshots, volumes or volume snapshots) created on the old cloud after today. If you do need to migrate data at a later date, you can use the OpenStack clients as described at https://www.chameleoncloud.org/docs/user-guides/openstack-kvm-user-guide/#d… As usual, don't hesitate to contact us via the help desk if you have any questions.

1 0

Chameleon is now publicly available!
by Pierre Riteau 28 Jul '15

28 Jul '15

We are happy to announce that Chameleon is now publicly available: new hardware with resource discovery, bare metal reconfiguration, metrics collection, and access to Infiniband and storage nodes is now available to all participating users! You can explore the Chameleon hardware by browsing through our resource discovery portal: https://www.chameleoncloud.org/user/discovery/. If you are already a Chameleon user, use the Bare Metal User Guide (https://www.chameleoncloud.org/docs/user-guides/bare-metal-user-guide/) to access the new hardware; if you are new to Chameleon get started in three simple steps at https://www.chameleoncloud.org/docs/getting-started/ Over the coming year we are planning multiple hardware and software upgrades. Among others, we will upgrade the hardware on our OpenStack KVM cloud, add to the storage options on the testbed, and, towards the end of the year, deploy new heterogeneous hardware units. On the software side, we plan to add more network management options, better snapshotting capabilities, and develop a library of Chameleon appliances. With the new public offering we are introducing a soft allocation model: all new Chameleon projects will be given a generous startup allocation that can be easily renewed or recharged as needs arise. The existing Chameleon projects will be given an unlimited allocation expiring at the end of October; users are encouraged to move to new projects. FutureGrid users can continue to port their FutureGrid accounts until the end of August.

1 0

The Chameleon Early User period is starting
by Pierre Riteau 17 Jun '15

17 Jun '15

We are happy to announce the beginning of the Chameleon Early User period. The Early User period has been created to allow for exercising the new hardware under realistic usage scenarios prior to public availability. Participants in the program will be able to gain unlimited access to new Chameleon hardware while Chameleon operators will be able to assess how the new hardware behaves under real-life conditions. Sign up for the Early User Program to get access: https://www.chameleoncloud.org/user/early-user-program/program/1/ Once approved you will be able to access the Early User Guide and follow its instructions to use the new Chameleon hardware.

1 0