Clondaq2

DELL Server

Model: RS720A-E12-RS24U

Serial number: R9S0MD0002LF

CPUs: 32 (2 AMD EPYC 9124 16-Core Processors)

Memory size: 386GB

/data disk space: 107TB SSD

old server

Clondaq2 is X4200 server running Solaris, s/n: 0626AM2523, f/n: F46985. It is second DAQ server serving mostly to run Event Recorder and handle SILO traffic.

Network ports:

clondaq2      - e1000g0
clondaq2-daq1 - ixgb0 (10GBit; was e1000g1 while 1GBit)
clondaq2-daq2 - e1000g2

After Solaris installation:

1. Create /ssa directory and make following links there:

cd /ssa
ln -s /mnt/raid0 active
ln -s /mnt/raid1 epmty
ln -s /mnt/raid2 postsilo1
ln -s /mnt/raid3 postsilo2

2. Create two files in /ssa directory:

/ssa/ENVIRONMENT contains:

set PARTITIONS = " raid0 raid1 raid2 raid3 "
set LOGFILE = "/ssa/logfile"
set CLON_LOG = "/usr/clas/logs"

/ssa/STUB.TARGET contains:

/mss/clas/dvcs2/data

(or whatever run period you are in)

3. Make /raid/stage_in link:

cd /
ln -s /ssa/active raid

4. Make /mss link:

cd /
ln -s /w/mss mss

5. Adjust Sudo config file.

6. Make sure ZFS RAID partitions are mounted.

7. Routing must be adjusted to make sure input and output traffics are taking different routes.

8. automounts settings must be modified for clondaq2: new file /etc/auto_stage must be created with following contents:

stage0 -rsize=1048576,wsize=1048576 sfs61:/export/stage
stage1 -rsize=1048576,wsize=1048576 sfs56:/export/stage
stage2 -rsize=1048576,wsize=1048576 sfs59:/export/stage
+auto.w.ep

and in file /etc/auto_master instead of /w auto.w.ep must use /w /etc/auto_stage.

NOTE: June 6, 2010 10GBit connection was lost (clondaq2-daq1), Bryan Hess checked it and found that unplugging/plugging helps; he replaced cable just in case, well watch it (I think it is not the first time ...)