Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.atnf.csiro.au/people/Tasso.Tzioumis/sched/AUTOTAPE.html
Дата изменения: Unknown Дата индексирования: Sun Apr 10 07:55:32 2016 Кодировка: Поисковые слова: jet |
AUTOTAPE is an obsolete parameter that only applied to tape. It can be ignored by users of disk-based recording systems (eg Mark5) and eVLBI.
AUTOTAPE is used to request that automatic tape allocation and control be used at stations that use the VLBA control system and have more than one tape drive. It has no effect on other stations such as those that use VEX or on VLBA controlled stations with only one tape drive. AUTOTAPE=1 or AUTOTAPE=2 causes the on-line system to determine the tape direction, head position and track allocation automatically. The schedules will only include instructions on whether to run, stop, or rewind. AUTOTAPE>=2 also activates automatic reversal of the tapes when the low tape mark is reached, even in the middle of scans; with AUTOTAPE=1 the tape will stop at the low tape point and wait for the next scan. The default action (any other value of AUTOTAPE including no value) is to have SCHED determine the tape motion and related information.
Automatic tape allocation is now required for most VLBA observations. AUTOTAPE=2 should therefore be specified in most schedules. AUTOTAPE=1 is risky and probably should not be used. It is provided only so that SCHED can create schedules for all modes that the on-line systems can handle.
Automatic tape allocation cannot be used for projects that will not be correlated on the VLBA correlator. SCHED will refuse to make such schedules. This relates to how the correlators learn how the tape was handled at the stations. Non-VLBA correlators generally obtain some of the tape positioning information from the schedules rather than the logs, and, with automatic allocation, the two can be very different.
When automatic tape allocation is specified (AUTOTAPE=2), any station that is using the automatic allocation will be removed from a scan if the source is below the hardware limits at both ends of the scan. One can tell when this happened because a ``D'' will appear next to the -- in the summary file elevation and azimuth for a scan where this happened. There is no override. If one is need, let Craig Walker know and it can be added. This function will mainly affect the Mauna Kea VLBA site which is often scheduled in scans before the source rises and which has long access times for the site techs doing tape changes.
For more information on how automatic tape handling works, see the AUTOMATIC TAPE ALLOCATION section.
The rest of this section concerns the use of AUTOTAPE for the Mark II system. It can be ignored by most (all?) current users.
For Mark II, AUTOTAPE requests automatic tape changes. Without it, tape changes must be explicitly specified. The value specified for AUTOTAPE, if there is one, should be a time. The default, if no argument is given, is 4:00:00, which is the recommended value. Specify a negative number to turn off tape changes and a positive number (a time) to change the default tape length. Tape changes are requested at intervals of the requested time beginning with the earliest start time of the whole schedule. If this would require a change in the middle of a scan, the change is moved to the start of the scan and all following changes are moved up. This is station dependent and so can cause non-simultaneous tape changes. The tape is assumed to be kept moving between scans.
The reference time for AUTOTAPE (Mark II use) can be altered with TPREF in the SCHED keyin file and with the TPTIME in the TAPEFILE. See Section 3.4 for information on TPTIME. TPREF is a time of the form hh:mm:ss. The automatic tape changes are requested at integral numbers of 4 hour intervals (for the recommended time) before or after TPREF. A date is not needed because 4 hours divides 24 hours evenly. Even if someone were to use the 2 or 6 hour VCR modes, those tape lengths still divide 24 hours evenly.
If TAPE is specified in a Mark II schedule for a scan while AUTOTAPE is set, a tape change will be requested and following tape changes will occur at 4 hour intervals from the time of the forced change. One specification of TAPE is usually all that is needed to get the tape changing sequence right if the earliest start time is a poor reference time.
For most Mark II projects, the default action of AUTOTAPE should be good. The major exceptions are projects that start on the half hour and tape changes are desired on the hour, and projects where it makes sense to have a short tape at the start on some stations to avoid short tape on other stations for which the source rises later. Both of these cases can be handled nicely with TPREF.
Note that for Mark II observations, if AUTOTAPE is turned off, no initial tape mount at the start of the project will be requested. This causes problems at some stations. It you insist on specifying all tape changes with TAPE, be sure to put one on the first scan for each station. This mode of setting tape changes is NOT recommended.