User Tools

Site Tools


operations:documentation.ivs.mark5_module_selection

Choosing the right module for the experiment

A module should be allocated by observatory staff, see the Experiment and Module Library.

As a rule R1 and R4 on a module by themselves (only at Hobart may R1 R4 go on a module together) and not mixed with other experiment types. If possible, please try to record an entire experiment onto a single module, most especially R4, CRDS, CRF, and R1 experiments. If it won't fit, please contact observatory staff - they will usually be able to put a module with enough space into the Mark 5.

If not indicated on the spreadsheet, you need to fit the experiment on available modules. Check to see how much data will be recorded; then see which modules are mounted in the Mark5.

Data requirement

On the Experiment and Module Library there should be a data size indicated for the current experiment; otherwise refer to a similar historical experiment with a data size. On this site the station staff also record which module has been provisionally allocated to the experiment and loaded in the Mk5.

Also at the end of the printed schedule summary is an estimate of the total amount of data that will be recorded for the experiment. For example, this experiment requires a tad under 1.4 TB:

 .
 .
 .
 203-1636b 6722 3C446     59 72 NEUTR  16:36:38  16:38:33    1:55  1354.0    *    
 203-1641  6739 0059+581  25  5 NEUTR  16:41:55  16:44:08    2:13  1357.7    *    
 203-1647  6756 1749+096 290 29 NEUTR  16:47:10  16:48:11    1:01  1362.0    *    
 203-1649  6773 0332-403 129 10 CCW    16:49:54  16:51:03    1:09  1363.9    *    
 203-1654  6790 1954-388 219 56 CCW    16:54:53  16:57:37    2:44  1366.1    *    

   Total  1371.4 Gbytes
   Total number of scans:   400

Module capacity checks

Typing mk5=bank_set? into eRemoteCtrl will return the VSNs of the modules currently loaded and their capacities in GB where the syntax is [Bank] [label]/[capacity]/[block size]. Here's an example output:

2014.198.18:30:05.89/mk5/!bank_set? 0 : A : USN-0087/2000/1024 : B : HOB+0051/6000/1024 ; 

So in this case Bank A is active (as it's listed first), it's label is USN-0087 and it has a capacity of 2000 GB. Bank B has HOB+0051 loaded and it's a 6000 GB module.

If the module that is in Mark5 already has some data on it, the amount of free space available can be seen in e-Remote Ctrl window “M5 Capacity” box above the pie diagram.

So which module should I use?

Rapid turnaround experiments, R1's and R4's should go on a module by themselves, ditto CRDS and CRF; do not double up with other experiments. As a rule of thumb, the experiment should go on the smallest module it will fit on available in the Mark 5. Multiple experiments of other types (AUG, AOV, OHG, T2 etc.) can be recorded onto large modules where space allows

Doubling up

All R4, experiments from Katherine and Yarragadee need to be directly shipped to Washington, so please avoid putting them on a module with another experiment.

Everything else needs to be shipped to Hobart for e-transfer to the correlator. Therefore, two experiments, if they fit, can go on the same module. However, R1's do need to get to Hobart ASAP. Often there will be an R1 straight after another experiment. It is undesirable, but in a pinch the R1 can go on the same module with the last experiment, providing the R1 is the LAST experiment to be recorded (and there is enough disk space!).

For example; there has been an Austral and there is still 3TB left on the module; there is no separate blank module to put the following R1 on; it is the middle of the night and there is no one at the observatory to load a blank module; the R1 will only use 2.5TB; then you may put the R1 on the same module after the Austral. In ordinary circumstances it would be preferred that the R1 go on a separate module.

/home/www/auscope/opswiki/data/pages/operations/documentation.ivs.mark5_module_selection.txt · Last modified: 2019/01/25 00:24 by Warren Hankey