[en] The Grid security mechanisms were designed under the assumption that users would submit their jobs directly to the Grid gatekeepers. Many groups are however starting to use pilot-based infrastructures, where users submit jobs to a centralized queue and are successively transferred to the Grid resources by the pilot infrastructure. While this approach greatly improves the user experience, it does introduce several security and policy issues, the more serious being the lack of system level protection between the users and the inability for Grid sites to apply fine grained authorization policies. One possible solution to the problem is provided by gLExec, a X.509 aware suexec derivative. By using gLExec, the pilot workflow becomes as secure as any traditional one

[en] In today's marketplace, the cost per Terabyte of disks with EIDE interfaces is about a third that of disks with SCSI. Hence, three times as many particle physics events could be put online with EIDE. The modern EIDE interface includes many of the performance features that appeared earlier in SCSI. EIDE bus speeds approach 33 Megabytes/s and need only be shared between two disks rather than seven disks. The interal I/O rate of very fast (and expensive) SCSI disks is only 50% greater than EIDE disks. Hence, two EIDE disks whose combined cost is much less than one very fast SCSI disk can actually give more data throughput due to the advantage of multiple spindles and head actuators. The authors explore the use of 12 and 16 Gigabyte EIDE disks with motherboard and PCI bus card interfaces on a number of operating systems and CPUs. These include Red Hat Linux and Windows 95/98 on a Pentium, MacOS and Apple's Rhapsody/NeXT/UNIX on a PowerPC, and Sun Solaris on a UltraSparc 10 workstation

[en] The Drift Scan Camera (DSC) System acquires image data from a CCD camera. The DSC is divided physically into two subsystems which are tightly coupled to each other. Functionality is split between these two subsystems: the front-end performs data acquisition while the host subsystem performs near real-time data analysis and control. Yet, through the use of backplane-based Remote Procedure Calls, the feel of one coherent system is preserved. Observers can control data acquisition, archiving to tape, and other functions from the host, but, the front-end can accept these same commands and operate independently. The DSC meets the needs for such robustness and cost-effective computing

[en] An approach to developing Online Systems for the VAX, under VMS, is described. A framework for integrating separate and diverse programs into a unified whole is described. Included in this is a scheme which permits multiple VAX processes to share the same terminal in an organized way. This makes use of a menu package, (MENCOM), which also permits a command line mode of operation, with dynamic switching between menu and command line mode. A single DISPLAY program can display the histograms of any program in the system, both the in-memory histograms and those previously stored on disk. A centralized message system is designed to handle all error and status messages. A general buffer scheme used to enter data from any input stream and to access data selectively is briefly described. This buffer scheme is covered in more detail, by D. Quarrie in his CDF Data Acquisition System paper, given at this conference

[en] Large scale QCD Monte Carlo calculations have typically been performed on either commercial supercomputers or specially built massively parallel computers such as Fermilab's ACPMAPS. Commodity clusters equipped with high performance networking equipment present an attractive alternative, achieving superior performance to price ratios and offering clear upgrade paths. The authors describe the construction and results to date of Fermilab's prototype production cluster, which consists of 80 dual Pentium III systems interconnected with Myrinet networking hardware. The authors describe software tools and techniques the t have developed for operating system installation and administration. The authors discuss software optimizations using the Pentium's built-in parallel computation facilities (SSE). Finally, the authors present short and long term plans for the construction of larger facilities

[en] The Online and Data Acquisition software groups at Fermi National Accelerator Laboratory have extended the VAXONLINE data acquisition package to include a VME based data path. The resulting environment, PAN-DA, provides a high throughput for logging, filtering, formatting and selecting events. It is discussed in this paper

[en] In this paper, the authors report on software developed in support of the Fermilab FASTBUS Smart Crate Controller. This software includes a full suite of diagnostics, support for FASTBUS Standard Routines, and extended software to allow communication over the RS-232 and Ethernet ports. The communication software supports remote procedure call execution form a host VAX or Unix system. The software supported on the FSCC forms part of the PAN-DA software system which supports the functions of front end readout controllers and event builders in multiprocessor, multilevel, distributed data acquisition systems

[en] The Grid security mechanisms were designed under the assumption that users would submit their jobs directly to the Grid gatekeepers. However, many groups are starting to use pilot-based infrastructures, where users submit jobs to a centralized queue and are successively transferred to the Grid resources by the pilot infrastructure. While this approach greatly improves the user experience, it does introduce several security and policy issues, the more serious being the lack of system level protection between the users and the inability for Grid sites to apply fine grained authorization policies. One possible solution to the problem is provided by gLExec, a X.509 aware suexec derivative. By using gLExec, the pilot workflow becomes as secure as any traditional one

[en] This paper reports on software developed in support of the Fermilab FASTBUS Smart Crate Controller. This software includes a full suite of diagnostics, support for FASTBUS Standard Routines, and extended software to allow communication over the RS-232 and Ethernet ports. The communication software supports remote procedure call execution from a host VAX or Unix system. The software supported on the FSCC forms part of the PAN-DA software system which supports the functions of front end readout controllers and event builders in multiprocessor, multilevel, distributed data acquisition systems

[en] A Distributed Monitoring System (NGOP) that will scale to the anticipated requirements for Run II computing has been under development at Fermilab. NGOP [1] provides a framework to create Monitoring Agents for monitoring the overall state of computers and software that are running on them. Several Monitoring Agents are available within NGOP that are capable of analyzing log files, and checking existence of system daemons, CPU and memory utilization, etc. NGOP also provides customisable graphical hierarchical representations of these monitored systems. NGOP is able to generate events when serious problems have occurred as well as raising alarms when potential problems have been detected. NGOP allows performing corrective actions or sending notifications NGOP provides persistent storage for collected events, alarms and actions. A first implementation of NGOP was recently deployed at Fermilab. This is a fully functional prototype that satisfies most of the existing requirements. For the time being the NGOP prototype is monitoring 512 nodes. During the first few months of running NGOP has proved to be a useful tool. Multiple problems such as node resets, offline CPUs, and dead system daemons have been detected. NGOP provided system administrators with information required for better system tuning and configuration. The current state of deployment and future steps to improve the prototype and to implement some new features will be presented