There are several sp_cop processes running, after increasing queue size:
SP_QUE_POST_SHMSIZE "268435456" # Post queue shared-memory Size
SP_QUE_Q_SHMSIZE "67108864" # Queue shared-memory Size
Some shareplex instances failed to start, "Not enough space calling shmget to get shared memory segment".
test2[tempdb][/home/spmgr] >ipcs -a
T ID KEY MODE OWNER GROUP CREATOR CGROUP NATTCH SEGSZ CPID LPID ATIME DTIME CTIME
Shared Memory:
m 3 0x00a5c581 --rw------- sfmdb users sfmdb users 8 10469376 1398 1414 17:15:53 17:15:53 17:15:53
m 5046276 0x000773f8 --rw-r--r-- spmgr spadmin spmgr spadmin 3 8388608 16072 17251 10:05:52 10:05:52 8:57:07
m 327685 0x7f5bd4d4 --rw-r----- oracle dba oracle dba 38 2197823488 9835 16350 9:59:21 9:59:21 17:28:18
m 1441798 0x001773f8 --rw-r--r-- spmgr spadmin spmgr spadmin 3 94208 16072 17251 10:05:52 no-entry 8:57:07
m 1114120 0x003773f8 --rw-r--r-- spmgr spadmin spmgr spadmin 3 67108864 16072 17251 8:57:08 10:05:52 8:57:07
m 786441 0x004773f8 --rw-r--r-- spmgr spadmin spmgr spadmin 1 67108864 16072 17251 8:57:47 10:05:52 8:57:07
m 819210 0x000b7e5d --rw-r--r-- spmgr spadmin spmgr spadmin 3 8388608 16729 12515 12:05:25 12:05:24 8:59:23
m 884747 0x001b7e5d --rw-r--r-- spmgr spadmin spmgr spadmin 3 94208 16729 12515 12:05:25 no-entry 8:59:23
m 1146892 0x003b7e5d --rw-r--r-- spmgr spadmin spmgr spadmin 2 536870912 16729 12515 8:59:25 12:05:24 8:59:23
m 786445 0x000773f7 --rw-r--r-- spmgr spadmin spmgr spadmin 4 8388608 16179 12478 12:05:10 12:05:10 8:57:59
m 917518 0x001773f7 --rw-r--r-- spmgr spadmin spmgr spadmin 4 94208 16179 12478 12:05:10 no-entry 8:57:59
m 753679 0x003773f7 --rw-r--r-- spmgr spadmin spmgr spadmin 3 134217728 16179 12478 8:58:00 12:05:10 8:57:59
m 262160 0x004773f7 --rw-r--r-- spmgr spadmin spmgr spadmin 2 134217728 16179 12478 8:59:40 12:05:10 8:57:59
m 262161 0x000bce06 --
Shared objects are mapped globally in hp-ux, total size of shared memory IPCs can't exceeds 1.75G.
According to "HP-UX 11i tuning and performance":
"Note that because shared objects are mapped globally, the total size of all shared objects in the system must fit within the third and fourth quadrants. Finally, the last 256 MB of the fourth quadrant is reserved for hard and soft physical address space (I/O space), which is used for addressing hardware devices based on slot number. This limits the fourth quadrant to only 756 MB of shared data."
Decreae queue size for all instances.
SP_QUE_POST_SHMSIZE "33554432" # Post queue shared-memory Size
SP_QUE_Q_SHMSIZE "33554432" # Queue shared-memory Size
