Xiao's 2011July run

From Monnier Group Research Wiki
Revision as of 03:38, 5 July 2011 by 192.153.157.220 (talk)
Jump to navigationJump to search

2011July4th

testing saving data. note: the dewar was pumping down, and hadn't been cooled down yet.

I ran RTSchedule and spooler under "xche@mirkwood:/MIRC/" with sudo, the camera setup was rows=16, cols=252, row off=5, col off=2, Nreads=5, Numberoffremes=1000, framesperreset=1000, which is the prism mode. I saved 65 files, and there were 8 files with frame countererror (missing one frame). The ratio was unusually high. (file No.: 0-64)

I changed the setup to rows=48, cols=252, row off=5, col off=2, Nreads=5, Numberoffremes=500, framesperreset=500, which is the grism mode. I saved 70 files, and only one file has frame countererror!! (file No.: 64-134)

changed back to prism mode. I saved 40 files and 6 of them has the frame countererror.

changed back to grism mode. saved 60 files and one of them has the frame countererror.

rebooted mirkwood, and using prism mode, still a lot of frame counter error. It tends to happen continuously, sometimes only received 0 frame out of 1000.

changed to grism mode, no error at all on spooler saving 45 files, but there is error on RTschedule just after saving data finished:

Timeout on blocking semaphore !!! numberCoherent: 10, numberOfCoadds: 15, savenum: 0 ERROR: Starting exposure failed -> 0x544F5554

I doubled the poolsize, but I can't even run the RTschedule.

I reduce the poolsize by half, the IRCam crashed. Also there were still frame counter errors. (I just notice that the UT time went into July 5th, is this the reason causing the crash?). I tested again, saved 70 files, no crashes, but >10 frame counter errors.

I restore the poolsize back. After restarting VME (VME was on), the situation seems better. I saved 100 files in prism mode, and only 5 of them have frame counter error, and all of them received 99X out of 1000 frames. And in grism mode, I saved 70 files, and one of them have the frame counter error.

Although these two modes both have frame counter error, but they are essentially different. In grism mode, when the error happens, it prints out the message length which is a very large number. But in prism mode, there is no such printouts.

Retrieved from "https://monnier.astro.lsa.umich.edu/research/index.php?title=Xiao%27s_2011July_run&oldid=2608"