Greetings,
I’m looking for a place to find info regarding Avante, which is built on SBclient / SB server (6.5.0), which in turn is built on something called Unidata 8.1 , which is running on a RHEL 7.6 VM that we just recently migrated from an HPUX machine… I really have no idea if this is the appropriate place to be asking questions about this… but I figured I’d throw it out here.
We have a user base of ~250-300 simultaneous users daily. Since migrating the stack to the RHEL box, we have noticed that anywhere between 3 to 10 or so users are locked out of their Avanté sessions. They may or may not actively log out; doesn’t seem to be any pattern. We have to ps -el processes every morning, find the orphaned processes, and kill -9 them. They are all running the process called UDT. After that, they can connect in fine.
Our troubleshooting efforts have us upgrading SBclient this weekend during maintenance; but we’re also looking into upgrading Unidata as well (one dot version up). While researching this, we noticed that our UDT configuration looks mighty … weird? here:
HP-Unix RedHat-Linux
LOCKFIFO=1 LOCKFIFO=0
SHM_MAX_SIZE=2147467264 SHM_MAX_SIZE=1073741824
SHM_GNTBLS=400 SHM_GNTBLS=40
SHM_LPINENTS=20 SHM_LPINENTS=10
MAX_REP_SHMSZ=0 MAX_REP_SHMSZ=2147467264
REP_LOG_PATH=/usr/igi/ud/replog REP_LOG_PATH=/usr/udthome/replog
First off, LOCKFIFO (an obscure word to Google, it seems), in all documentation I could dig up so far indicates it should definitely be set to 1. However, I cannot find any explanation as to what it is or what it means. I have not had a chance to research the other settings yet.
Lastly, if this is completely irrelevant to what y’all discuss here in these forums, and you have an idea who WOULD know this information, feel free to let me know! Thanks!!
“Goonies never say die!”