more draft

docs/cheatsheet.md

@@ -1,6 +1,6 @@
 # LMTOY cheat sheet
 
-A brief reminder on the commands we use to operate the pipeline and everything around it
+A brief reminder on the commands various stakeholders use to operate the pipeline and archive ingestion.
 
 ## 1. Lightweight TAPs Pipeline: lmtslr@malt
 
@@ -9,30 +9,62 @@ taken. The lightweight TAP's are sent to Unity, where they can be viewed.
 The user **lmtslr** runs the commands here.
 
 
-1. Start the data catcher
+1. When new software in LMTOY is available, use
+
+       lmtoy pull
+
+   and pay attention if anything needs to be recompiled. Most is python, and doesn't need any
+   further action. Only the gridder may need a recompilation, but this is rare:
+
+       cd $LMTOY
+       make update
+
+2. Start the data catcher
 
        cd ~/SLpipeline.d
        SLpipeline_run.sh > SLpipeline.log 2>&1
        tail -f SLpipeline.log
 
-2. Summary of what rsync does
+   A file **SLpipeline.pid** contains the **process id** of **SLpipeine_run.sh**.  This
+   needs to be killed before running another one (e.g. if software was updated)
+
+       kill -9 $(cat SLpipeline.pid)
+       rm SLpipeline.pid
+
+2. This data catcher will allow some "permanent" extra parameters (e.g. pix_list=-13) to
+   the pipeline, if added to the file **SLpipeline.in**:
+
+       exit SLpipeline.in
+
+3. Summary of what rsync has done is a nice way to remotely keep track of progress
 
        tail -f rsync.log
 
-3. Re-execute a pipeline with improved parameters (only for urgency)
+4. Re-execute a pipeline with extra parameters (only for urgency)
 
        SLpipeline_run1.sh 113271 2024-S1-MX-24 dv=10 dw=10
 
+   this is a convenient way to fix really bad pipeline runs.
+
+5. After this the most recent 100 TAP's produced are on this list:
+
+       http://taps.lmtgtm.org/lmtslr/lmtoy_run/last100.html
+
+6. There is also a way to make these TAPs available via a tunnel on a browser that runs
+   on **malt** itself, but this is for emergency only and is usually not available as the TAP
+   copies to Unity seem pretty reliable now.
+
 ## 2. Script Generator
 
-This work can be done anywhere, since it's git controlled. But it depends on the user having
-installed the **gh** command, otherwise it's annoying work in the browser
+This work can be done anywhere, since it's git controlled. But the initial bootstrapping
+depends on the user having
+installed the **gh** (github CLI) command, otherwise it's annoying work in the browser.
 
-0. Set the project id in a convenient shell variable we are working with in this cheat sheet
+0. Set the ProjectId in a convenient shell variable **$PID** we are working with in this cheat sheet
 
        PID=2024-S1-MX-24
 
-1. Bootstrap a new project script generator (the **gh** command is needed here)
+1. Bootstrap a new script generator for this project (the **gh** command is needed here)
 
        cd $WORK_LMT/lmtoy_run
        ./mk_project.sh $PID
@@ -41,15 +73,15 @@ installed the **gh** command, otherwise it's annoying work in the browser
    The last step, the links, are convenient for going back and forth between where the pipeline
    data are and where the script generator is.
 
-2. Edit Makefile to add lmtoy_2024-S1-MX-24 to the specific year
+2. Edit **Makefile** to add lmtoy_2024-S1-MX-24 to the specific year
 
        edit Makefile
 
-3. Edit comments.txt to add 2024-S1-MX-24
+3. Edit **comments.txt** to add 2024-S1-MX-24
 
        edit comments.txt
 
-4. Edit REDAME.md to say a few interesting things, make sure $PID is correct
+4. Edit **README.md** to say a few interesting things, make sure $PID is correct
 
        edit README.md
 
@@ -58,7 +90,7 @@ installed the **gh** command, otherwise it's annoying work in the browser
        git commit -m "new project"  Makefile comments.txt README.md
        git push
 
-5. Typically at the end of observing season, a record of all the obsnums is saved as well:
+5. Typically at the end of observing season, a record of all the obsnums is saved as well in **lmtinfo.txt**:
 
        lmtinfo.py $PID > lmtinfo.txt
        git add lmtinfo.txt
@@ -129,12 +161,24 @@ installed the **gh** command, otherwise it's annoying work in the browser
 
         make summary index
 
+   The following URL's will then give access to this project:
+
+	xdg-open http://taps.lmtgtm.org/lmtslr/$PID
+
 8. Ingest in the archive. Here we have to make sure that the final runs were done with "admit=1 sdfits=1 srdp=1"
 
-cd $WORK_LMT/lmtoy_run/lmtoy_$PID
+        cd $WORK_LMT/lmtoy_run/lmtoy_$PID
+        find_obsnum_and_ingest.sh ?args?
 
+   Details on the arguments to be determined. In theory, the script generator knows what to ingest, but perhaps
+   these need to be confirmed on the commandline. Once this button is pushed, there's no way back.
 
 ## 4. Helpdesk Pipeline: lmthelpdesk_umass_edu@@unity 
 
 Many things overlap with how the final pipeline is run and submit, except the work in git. Generally the DA's work
 in a git branch, so the work can be shared with the main pipeline work. More details to come here.
+
+
+## 5. PI remote execution
+
+Something needs to be written down how a remote PI webrun is impacted when new data arrives.