SSH tunnel for github on jlab clon*** machines:
(First make sure you have an RSA key pair in your clon home that is registered with your github account)
SSH to the clon machine as usual through the gateway
> ssh user@hallgw.jlab.org -t ssh user@clon***
Then create a tunnel to github.com:22 through jlabl4
> ssh user@jlabl4 -L XXXX:github.com:22
You now have a tunnel from the inside host port XXXX to github.com port 22.
You then need to open a new terminal window and ssh to the same clon machine through the gateway
> ssh user@hallgw.jlab.org -t ssh user@clon***
Now in that terminal, you can access github like this:
> git clone ssh://git@localhost:XXXX/slaclab/heavy-photon-daq.git
When you first clone a repo, you'll need to mod the submodule config. Edit .gitmodules:
[submodule "drivers"]
path = drivers
url = ssh://git@localhost:5555/slaclab/aes-stream-drivers.git
branch = master
> git submodule init
> git submodule update
And avoid checking the submodule path change into the shared repository.
L6 Testing Preparation
Preparation
- Prepare laptop with 1 Gb network card and both devboard DAQ and QA software (Omar)
- Locate devboard we will use at JLab (Omar)
- Use laptop to test communication with devboard
- Locate L4-6 hybrid adapter board (samtec to DB44)
- Pelle has possession of one and Omar has another. Both look fine.
- Build new module cable (Pelle)
- Build new power/signal cables if needed (Pelle)
- Test laptop using existing Group C setup and with new module cable (Omar/Matt)
- Locate power supplies at JLab, check with Stepan (Pelle)
- Need 2 power supplies to provide 1.25 V and 2.5 V to hybrid
- Need 1 power supply to provide ~5V
- Need single supply (Keithly) to provide bias up to 180 V (possible higher)
- Find spare 30m fiber (Pelle)
- Find extra transceivers (Ben)
Hardware needed at JLab
- Laptop (obvious ...)
- Optical to cat 6 (ethernet) converter
- LC-LC fiber
- Cat 6 cable (ethernet)
- 30m quad fiber spare (need to be a cross over fiber)
FEB Testing Preparation
Requires:
- Laptop/desktop in Hall with vnc to the main DAQ
Link status:
- Link 6 is bad
- Link 10 is unused
- Link X is the bad one (find out)
- Link 11 carries FEB6
Guide steps for FEB link debugging
- Find and understand DAQ map
- Reseat fibers/transceivers on flange and RTM on the faulty FEB
- Switch fibers on the flange and on the RTM
- If error follow RTM -> swap in new transceiver
- If error follows fiber -> swap fiber
- Swap transceiver on the flange
- If all fails through 4:
- Test new firmware images for the bad FEB that has different GTX configurations
- Consider swapping the mSAS of the faulty FEB to a different channel on the inside of the flange.
- Take the FEB plate out
FEB Firmware Testing
- Install boot loader to FEBs
- Load the same firmware image we are using now
- Verify with just link and configuration
- Maybe take data with CODA
- Assume successful from 2.
- Install new firmware with latest libraries
- Verify that it's backward compatible with software
- Debug if necessary, this will need to work before moving on (REQUIRES git, RCE compilation, etc.)
SVT Removal Preparation
Supplies
- Prepare data loggers for shipping
- Order desiccant (Stepan to order)
- Gloves (M and L)
- Arm covers
- alcohol wipes
- zip ties
- VCR caps
- VCR gaskets
- Have Lupe ship box to JLab
Procedure