|
|
New VS FAQ
|
Quick links within this page:
The New VS is a virtualization, an emulation if you like, of the Wang VS machine. It runs on modern Linux servers and is free of proprietary Wang hardware unless you opt to use the available "928" card to support legacy Wang periperals such as workstations, printers and telecomunication gear. Most New VS sites do not use the 928 option. Wang workstations and printers are mostly replaced by Lightspeed features. Wang telecommunication still has some uses, namely to support SNA, WSN to legacy systems, X.25, and LU 6.2.
Since the New VS is a true and faithful virtualization of the Wang VS, it runs the VS Operating System, all VS compilers, all VS utilities, and all VS application software, all unchanged.
The New VS:
Realistic "futures." The New VS will have:
The New VS has very important implications for Wang VS sites:
Even without counting on any new features not yet available, the New VS immediately provides you with new capabilities simply because of its virtualization technology:
The New VS seamlessly replaces all models of Wang VS and offers performance far exceeding any legacy Wang VS.
Yes. Not only does the New VS support VS clustering, for the first time it has the ability to cluster up to 16 VS systems, something that was never fully implemented in the legacy VS OS.
Unlike legacy RSF, which utilized FDDI fiber at 100 Mb/sec, the New VS RSF utilizes IP networking, typically at Gigabit speed. Thus the New VS communicates potentially ten times faster than the legacy Wang VS. RSF may be bound to the secondary network interface to isolate it from normal network traffic and maximize RSF throughput. In a two-node VS cluster the two network interfaces would be directly connected to each other. In a cluster larger than two nodes the systems would be connected to a Gigabit switch.
Yes, the New VS uses RAID for all its storage. This is transparent to the VS OS and applications and serves to increase the fault tolerance of the system. No New VS system is placed in the field without at least internal RAID.
The New VS stores VS disk volumes as Linux files containing byte-for-byte images of VS disk drives. The image files are themselves contained in RAID arrays and thus protected from simple physical disk failures. Multiple virtual VS volumes are stored in RAID arrays. There is no correspondence between virtual VS volumes and the physical disk drives of the New VS. Virtual VS volumes may be stored in any combination of internal RAID or external enteprise storage.
GENEDIT remains as the principle configuration tool, unchanged from how the legacy VS is configured. What is different is how what the VS thinks are physical devices are "connected" to the virtual VS. Whereas cables and ID settings were used in the legacy VS to connect physical devices, these connections are done in software in the New VS. In the System Control Unit (SCU) there are configuration pages similar to GENEDIT in which the virtual VS is told where to find the devices configured in GENEDIT. Thus a disk, for instance, configured in GENEDIT to be on IOC 3 in slot 2 would be configured in the SCU to point to a virtual disk image file in IOC 3, slot 2. A physical tape drive, for instance, configured in GENEDIT to be on IOC 4 in slot 5 would be configured in the SCU to point to the actual Linux SCSI tape drive device in IOC 4, slot 5.
Thus there is a layer of configuration that replaces the physical plugging and cabling that "connects" devices in the legacy Wang VS and provides the connectivity to the virtual VS so it knows where the devices are that are called for in the VS GENEDIT configuration.
The New VS can configure up to a 34.3 GB VS disk volume, the largest supported by VS OS 7.54.12. Note that using such a large volume on the legacy Wang VS was problematic because the legacy VS SCSI subsystem still operated at SCSI-1 speed of 5 MB/sec. Thus it could take a very long time to perform any large disk operations on the legacy VS while the New VS can perform large disk operations with ease and efficiency. Note also that in the VS world of very efficient disk storage formats and file compression, a 34.3 GB disk is truly huge.
Yes. In fact, the New VS can be used for VS systems larger than any that exist in the legacy VS world. The top tier of New VS performance runs at nearly three times the performance of the legacy high-end VS18950 and provides much faster disk file I/O than any legacy VS as well as eliminating the bottleneck of Lightspeed gateway PCs and their WLOC or SCSI cards. Not only can you replace the largest Wang VS systems in existence, you can have a much better performing system with faster response and greater fault tolerance.
There are production New VS systems in operation that have at least 500 users. The limit of users is presently prescribed by the VS OS and by Lightspeed. The VS presently has a limit of 999 devices or users, and Lightspeed cuts that in half by utilizing an invoked background task for every foreground user task. Thus it is possible to configure nearly 1,000 non-Lightspeed users on a New VS or nearly 500 Lightspeed users. All this will change with the upcoming release of VS Operating System 7.54.20, which will double the limit of devices/users to 1,999.
Yes. Because virtual VS volumes are simple Linux files, they may be stored on any device that functions in Linux as a file system. The most reliable form of storage, though, is in the RAID that is internal to all New VS systems. The internal RAID is always available, unlike some external storage subsystems. Backups, however, may certainly be sent to external enterprise storage. Special volumes and work volumes may be configured on external storage for various reasons.
You will be delighted. We normally configure the physical server to have much more disk space than the original legacy Wang VS. In some cases this is unavoidable because modern servers cannot be configured with less than about 73 GB of disk, which is more than most VS systems have had. In other cases we simply make sure that the New VS will have much more space than the original VS. Since virtual SCSI IOCs and virtual disks can be created and configured at will, this means that from initial setup through day to day operation it will be easy to make sure you have sufficient VS volumes to get your work done. It will even be easy to create and configure virtual VS volumes for temporary use, destroying them when no longer needed.
The relatively cheap availability of disk space in modern servers changes the entire picture of disk space in the VS world. While previously an increase in disk space required a purchase authorization, now it requires only the decision to create and configure the additional space, as long as the physical server has sufficient free space available, and we strive to configure all New VS systems to have more than sufficient disk space.
More than you will likely need. Since the New VS emulates the Wang VS machine, and since the CP types it emulates generally have 15 I/O slots, and since at least one I/O slot is needed for workstation connectivity, up to 14 slots are avaiable for virtual SCSI IOCs, each supporting up to 14 devices. So the maximum number of virtual disk drives configurable is 196, and each virtual disk drive can be as large as 34.3 GB. The maximum storage capacity would therefore be 6,722 GB, or 6.7 terabytes. Since the New VS is virtual, it's possible that the number of supported IOC slots may be increased in the future, but for now it seems sufficient.
Presently up to about 500. The capacity depends on the VS OS and on the performance tier you select. The VS OS presently has a limit of 999 devices/users, and if you use Lightspeed the capacity is half that because Lightspeed invokes a background task for every foreground user task. So at the highest performance tier you could presently have close to 1,000 non-Lightspeed users or close to 500 Lightspeed users. The upcoming new VS OS, 7.54.20, will double the capacity, allowing close to 2,000 non- Lightspeed users or close to 1,000 Lightspeed users.
We offer a non-Lightspeed workstation emulator, bundled with the New VS. It runs in Windows and in Linux and requires no install, no Registry modification. It is not as robust as Lightspeed but works perfectly well for logging on and running applications. Also, it does not do file transfer or document conversion. It is simply a workstation emulator.
One of our customers has also developed a Visual Basic workstation emulator that looks very much like Lightspeed. We believe this will also be available to our customers.
You could also run legacy VS workstations, but we don't recommend that. The legacy workstations require our 928 card and additional external legacy equipment to connect to the New VS, and are a maintenance problem, as workstations and spare parts are becoming scarce.
Yes. The New VS uses IP networking for WSN links with other New VS systems, potentially communicating more than ten times faster than legacy systems were able to communicate. The New VS also has the ability to operate legacy WSN links with legacy VS systems by means of the 928 card and external legacy TCB equipment at legacy speeds.
Yes. Support for VS TCP/IP is comprehensive. The FTP server works, as do programmed VS TCP/IP services. VS TCP/IP works in the New VS exactly as it does in the legacy Wang VS.
Yes. VS programmed async is supported using LAN-based MOXA devices transparently to the VS and VS applications. Async ports are configured as TCB1/Async, each one bound to a MOXA RS-232 async device. No changes are required to async programming.
Yes. All forms of transfer into and out of the New VS are faster than they were in the legacy VS. Workstation I/O is faster, File Exchange transfers are faster, WSN file transfers are faster, and VS TCP/IP file transfers are faster. In addition there are new ways of transferring data, and those are much faster than traditional VS data transfers.
Yes. We have added functionality to provide for transferring file data between the VS and Linux environments.
Completely. The New VS is a 100% seamlessly binary compatible emulation of the Wang VS machine. It runs everything with no changes. This means that there is no "conversion," no "migration," simply loading the New VS from the disk volumes or backup tapes of the legacy VS. The first onsite install we did was installed, accepted and put into production in less than one elapsed day. There is no conversion on Planet Earth that can claim to be able to do this. Most of our installs take a few days, which means that code or data freezes are very short and manageable.
There are somewhat different solutions for the different tape technologies, and different solutions for short and long term. We suggest that you convert your tapes to virtual tapes so they will forever be readable. We may be able to help convert your old tapes.
In all cases, if you have a large number of library tapes that must be preserved, you can either convert the tapes yourself or you can obtain outside help to convert them. A shortcut is to be able to read them on the New VS, in which case they can be converted directly to virtual tape images using our Integrated Virtual Tape feature. Once converted to virtual tapes the images will remain readable essentially forever, without tape errors.
Almost too many! The most exciting New VS feature in support of backups is Integrated Virtual Tape. This feature allows you to write backups to virtual tape image files in a library. Tape image files, once created, no longer suffer from tape read errors, so your tapes remain readable essentially forever.
The Integrared Virtual Tape feature provides for virtual tape drives to be configured, and to redirect input and output VS tape operations to/from virtual tape image files. Files in the virtual tape library are identified by VS tape volume ID. Thus, any present system of globally unique tape IDs can be duplicated in the virtual tape library. Futher, each virtual tape drive can optionally utilize a different library, allowing you to maintain multiple sets of globally unique tape volume IDs. To put VS backups in context, you will continue to be responsible for backing up individual VS volumes because no form of system-wide backup can provide for restoring individual VS volumes, libraries or files. You will not, however, have to back up to physical tape on the New VS. You will probably make backups to virtual tape, stored either in the RAID in the New VS or in external storage. You will have many options, though:
No. All your programs will run without any changes.
Yes. Everything works exactly the same on the New VS as it does on the legacy Wang VS. You might, of course, end up with newer versions of compilers and utilities. Upgrade of your licensed VS software products is part of the New VS install.
The VS OS and all system software continues to be supported and new versions released from time to time. The VS software in your New VS will be upgradable just as the software in your legacy Wang VS was always upgradable.
The New VS is the official new generation of Wang VS. It is sanctioned and certified by the VS group in Compucom in Tewksbury, Massachusetts, home of the VS.
Yes. PACE works exactly the same on the New VS as it does on the legacy Wang VS.
Your system will continue to be secure after you move to the New VS. The New VS uses Linux merely as the basis of an appliance. VS users do not log into Linux and the system generally runs no servers. The system runs just one application -- the virtual VS -- and that application offers no general services to the network.
We generally use Dell PowerEdge servers although we are working to qualify IBM xSeries servers as well. A typical configuration consists of:
More processor cores do not necessarily help. The virtual VS uses one processor core and its I/O threads use another. Additional processor cores are essentially wasted and will not contribute to better or more efficient performance of the virtual VS. The main reason for using a Quad Core is that it may be the only type of processor available that offers the required speed. The industry is moving in the direction of more cores.
The memory must be mirrored, with usable memory no less than 2 GB. Each server has a specification of how many DIMM memory modules must be installed in order to enable memory mirroring. If that number results in having more than 2 GB of mirrored memory, that is OK.
Generally, more memory is better and will increase the file buffering that Linux does, thus reducing the latency of disk file operations. It's also the case that very large systems may call for more memory than smaller systems.
This is an integrated feature of the Dell PowerEdge. It is not a standard plug-in card. It is vitally important to have battery bacup for the RAID cache, otherwise disk corruption could occur if power is removed from the server.
Remote access is part of the minimum specification of the server. This is how the software is installed in cases of overseas sites where onsite installation is not done. It is also how support is provided in cases where Linux is damaged or otherwise not behaving correctly. Constant remote access is not required; but it must be available during install and if and when support is required.
Although physical tape is being used less and less, we still usually configure a tape drive on most New VS systems, especially where the ability to read existing tapes exists.
The removable disk drive bay is used for "bare metal" backup and restore and optionally for backups that may easily be removed from the system for off-site storage. It is also used to clone the newly installed production system in order to make a backup or DR system. It can further be used to transport replicated volumes from the production system to the backup or DR system.
The default tape drive for New VS systems is the DAT72, which has a capacity of 36 GB uncompressed and up to 72 GB compressed. It may be used for straightforward VS volume backups to tape, may be used for "bare metal" backup and restore, and may be used to read DDS-3 or DDS-4 or DAT72 tapes containing archival data or data being moved from the legacy Wang VS.
Large systems with a large RAID array may call for an LTO3 tape drive, which has a capacity of 400 GB uncompressed and up to 800 GB compressed. It may be used in all the same ways as the DAT72 but will use fewer tapes for large operations, due to its greater capacity.
Serial Attach SCSI (SAS) drives are the most suitable for the New VS. We do not use SATA drives.
The smallest New VS systems have a single RAID1 (mirror) array containing Linux, the virtual VS software, and all the virtual VS volume images. Larger systems may have an additional RAID1 array, and the largest systems will have a RAID5 array in addition to the base RAID1 array. The base RAID1 array is as small as possible, to facilitate "bare metal" backup and restore. If there are multiple RAID arrays the base RAID1 will contain Linux, the virtual VS software, and one or more IPL volume images, with additional VS volume images located in the secondary array. This provides for bare metal restore to put in place the files necessary to boot Linux and to IPL the virtual VS.
Extreme reliability requirements may call for storing the VS volumes in an external RAID with dual controllers and access from dual servers, eliminating single points of failure. This is not normally required.
Some customers wish to have a duplicate backup or DR system to decrease the Time to Recovery in the event of a serious failure. Such a duplicate system must be ordered at the same time as the production system in order to ensure that the two machines are in fact the same, with identical motheboards and chipsets. The server offerings change frequently, and it may be impossible to obtain a truly duplicate machine at a later time.
Uninterrupted power (via a UPS) is vital to a system that performs significant memory buffering and caching of data. Site UPS is not sufficient to provide the highest level of uninterrupted power for two reasons: First, site UPSs can and do fail. Second, the power path between a site UPS and the server is likely to be long and subject to interruption. A small, dedicated UPS located immediately adjacent to the server is the best measure to ensure the highest level of uninterrupted power.
Yes. Onsite installation is available in the United States and around the world. It is an extra-cost option and is based on sending a technician from one of our U.S. facilities to the installation site.
For U.S. customers we purchase the server and prepare it at one of our facilities in the U.S., then reship it to the customer site. Onsite install is optional. For overseas customers the customer or his agent procures the server to our specifications and we either remotely install the software or perform an onsite install.
The New VS is exclusively sold by TransVirtual Systems. This is as a result of the termination of legacy Wang VS support and appointment of TransVirtual Systems as the exclusive reseller of the technology by Getronics in July, 2008. Additionally, in August, 2008, Getronics North America was sold by the owner, KPN, to Compucom, a U.S.-based IT firm, thus severing all connections between Getronics and the VS.
The USI utilities, Back Burner, Over Easy with Eggshell and Short Order, are relicensed for the New VS by Compucom.
Lightspeed is licensed by Lightspeed NVS, part of the TransVirtual family of companies. There is one authorized reseller in Sweden, serving the European community.
TransVirtual provides first-level support for the New VS and in-depth support for Linux and the virtual VS software. Compucom provides support for the VS Operating System and all optional licensed VS software products that do not originate with third parties.
The USI utilities, Back Burner, Over Easy with Eggshell and Short Order, continue to be supported by Compucom.
Lightspeed is supported by Lightspeed NVS, part of the TransVirtual family of companies. There is one authorized reseller in Sweden, serving the European community, authorized to provide Lightspeed support. Getronics is not an authorized reseller of Lightspeed and is not authorized to provide Lightspeed support, updates or upgrades.
Good luck. Support for the legacy VS was discontinued in July, 2008. With the exception of some very pricey ongoing maintenance contracts issued out of Tewksbury there is no official support for the legacy Wang VS, and supported parts are not generally available. Stocks of spare parts in the hands of third party maintainers are mostly old, not having been manufactured in the last 10-20 years. Although the legacy Wang VS has enjoyed an exceptional record of reliability, everything eventually gets old and begins to malfunction. So it is with the legacy VS.
The New VS is a no-brainer. It replaces any legacy VS quickly efficiently, with no risk. In contrast, VS conversions have an awful track record of high cost and risk of failure. Even when nominally successful, VS conversions have almost always fallen short of the functionality and ease of use of the VS systems they replaced. Some have crashed and burned. It is not unusual that VS customers attempting conversions have spent millions to replace a system with a value of only low six figures.
Anyone can claim to be able to maintain anything. The fact is that since the sale of Getronics North America by KPN to Compucom in August, 2008, remaining Getronics entities worldwide have had no more connection to the VS or to the source of supported parts and software, which come from the VS group in Tewksbury, Massachusetts. Various Getronics entities no doubt have some stocks of spare parts, but most of those parts were manufactured 10-20 years ago and are as old as the machines being maintained. Getronics is now just another third party maintainer. They had the better part of five years in which to have promoted and sold the New VS, but they didn't do it. Now they want to keep customers on an aging, failing platform in order to make money at the expense of customers making the wise decision to upgrade to the New VS. In the end only those customers can decide what is the intelligent thing to do, but they should do it in the light of full information.
If you want to utterly destroy all the features and benefits of using your VS, sure, go ahead and convert your system to Oracle. It will only cost you a very expensive Oracle license, the price of a beefy Oracle server, and breathtaking fees for the conversion itself. What you end up with will not be a VS, will not look like a VS, and will not behave like a VS. You will have to retrain your people and pay for one or more database administrators. You will probably not be able to retrain your programmers and will have to replace them with database programmers. The conversion will take considerable calendar time and require risky project management to have any chance of success. All of that when you could upgrade to a New VS in a matter of mere days, with no retraining, retaining full VS functionality, and zero risk of failure.
We have been watching VS customers do various kinds of conversions in the last 20+ years. We know of no completely successful conversions but we know of many partially or completely failed conversions. In all cases, enormous amounts of money have been wasted on VS conversions. Before the New VS, the alternative was to upgrade to a more powerful model of VS at a completely predictable cost and with no risk of failure. Now that the New VS is the current generation of VS, the alternative is to upgrade to the New VS at a completely predictable cost, with no risk of failure. In both cases the upgrade involves essentially no cost in calendar time, as it takes place the same day or within a few days.
In more than 20 years we have never heard a very good justification for any VS conversion but we have witnessed many VS customers spending many times what a simple upgrade to a more powerful model of VS would have cost. In the worst cases we have seen VS customers spend $10-20 million to replace a VS, and in some cases the conversions have crashed and burned, the money utterly wasted.
If a customer is bound and determined to listen to the wrong advice and embark on a costly conversion, all we can say is, "Knock yourself out." Have a ball throwing your money away. Meanwhile we will be upgrading intelligent customers to the New VS in mere days, retaining full VS functionality, with zero conversion and zero retraining. The difference between a conversion and upgrading to the New VS is like the different between night and day. The New VS is truly a no-brainer.