HP 3PAR StoreServ 7400 4-node HP 3PAR OS 3.1.2 Release Notes - Page 9

Description, Issue ID, Virtual Volume.

Get HP 3PAR StoreServ 7400 4-node PDF manuals and user guides View all HP 3PAR StoreServ 7400 4-node manuals
Add to My Manuals
Save this manual to your list of manuals

Page 9 highlights

Issue ID Item Description 54956 2.3.1 MU1: sysmgr panicked, system When resynchronizing a volume over 2TB in size to a rcopy secondary pending pfail recovery system, a 32bit counter used to represent the number of contiguous blocks which has ""delta"" data since the last resync may overflow. HP 3PAR OS has since been enhanced to use 64bit counters. 55196 Configure fabric ports in target mode Addresses an issue in earlier software releases where Emulex HBA after node replacement ports configured in host/fabric mode could be switched to initiator mode after node replacement. 56781 Allows IMC support for scheduler In 3.1.1.MU1, there will be IMC support for the scheduler because this bugfix enables the required functionality in the HP 3PAR OS. 57495 sysmgr panic and degraded single The command setvv -snp_cpg (remove the snapspace CPG) was esi port for all cages when removing causing the System Manager to crash for Virtual Volumes created via vv from cpg createaldvv, which had space reclaimed. This command should now work without error, provided that there is no snapshot on the Virtual Volume. 58251 Remote copy traffic may stall if destination link IP address corresponds to a local link address To avoid potential problems with stalled remote copy traffic, in 2.3.1.MU5 the creatercopytarget and admitrcopylink commands will not permit destination link IP addresses that correspond to local RCIP link addresses or the address of the management interface. 60029 Fixed chunklet relocation after media Media failures on certain chunklets would sometimes prevent relocation failures. of other chunklets from the raid set as well as returning a preserved LD back to normal. This has now been fixed. 60141 HP 3PAR system Panic - tpd: Assertion point: file: cmp.c, line: 16883 When a controller node rejoins cluster, HP 3PAR OS attempts to rebalance the load across system. During the rebalancing process, the current HP 3PAR OS does not handle the out of resource condition which can cause entire system to be offline. The fix is to make sure there is enough resource before starting the load rebalancing process. 60272 Failed snapshots for atlantis cluster - There is a new database created in the "Atlantis" instance, called "OAuth". Notice that there is a "space" in the name of the database. For some reason, definitely a bug with the SQL writer, the writer fails to come up; which explains why we don't see it in the list of vssadmin writers. This topic has been discussed, and the user apparently found the solution with no acknowledgement from Microsoft. http:// social.technet.microsoft.com/Forums/en-US/dpmsqlbackup/thread/ 432b4e38-6c7c-4fe0-abe5-769275784fcf/ 60494 When deleting a very large and If we have a heavily loaded system and have snapshots with lots of filled snapshot can lock up a CPU for data, the meta data tables which are heavily populated, then on long time deleting some of these snapshots the meta data tables are traversed causing the cpu on occasions to be held for an excessive amount of time. This can cause the locking out other processes/tasks from running and can cause the background internode comms processing to be delayed resulting in the erroneous belief that one of the other nodes has died. This fix prevents the cpu from being held for an excessive amount of time while traversing the the meta data tables during snapshot removal. This fix includes an enhancement (65646) to cover the case where the meta data tables are being merged during snapshot removal. This fix with enhancement is applied to 2.3.1.MU5, 3.1.1.MU1, 3.1.2. The fix without the enhancement is in 3.1.1.(This is also logged as # 52911). 61058 Raw space alerts Excessive raw space alerts have been seen when AO is configured on a system. In 3.1.2 we have provided the ability to set raw space alert threshold to 10g to suppress these alerts. 61363 Removal of volume from exported VV It was possible to try to remove a VV that had been just removed from set could cause sysmgr crash an exported VV set before the VV export was fully cleaned up. In some instances this would cause sysmgr to crash and restart. Modifications to the OS 9

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
  • 37
  • 38
  • 39
  • 40
  • 41
  • 42
  • 43
  • 44
  • 45
  • 46
  • 47
  • 48
  • 49
  • 50
  • 51
  • 52
  • 53
  • 54
  • 55
  • 56
  • 57
  • 58
  • 59
Description
Item
Issue ID
When resynchronizing a volume over 2TB in size to a rcopy secondary
system, a 32bit counter used to represent the number of contiguous
2.3.1 MU1: sysmgr panicked, system
pending pfail recovery
54956
blocks which has ""delta"" data since the last resync may overflow.
HP 3PAR OS has since been enhanced to use 64bit counters.
Addresses an issue in earlier software releases where Emulex HBA
ports configured in host/fabric mode could be switched to initiator
mode after node replacement.
Configure fabric ports in target mode
after node replacement
55196
In 3.1.1.MU1, there will be IMC support for the scheduler because
this bugfix enables the required functionality in the HP 3PAR OS.
Allows IMC support for scheduler
56781
The command
setvv -snp_cpg
(remove the snapspace CPG) was
causing the System Manager to crash for Virtual Volumes created via
sysmgr panic and degraded single
esi port for all cages when removing
vv from cpg
57495
createaldvv
, which had space reclaimed. This command should
now work without error, provided that there is no snapshot on the
Virtual Volume.
To avoid potential problems with stalled remote copy traffic, in
2.3.1.MU5 the
creatercopytarget
and
admitrcopylink
Remote copy traffic may stall if
destination link IP address
corresponds to a local link address
58251
commands will not permit destination link IP addresses that correspond
to local RCIP link addresses or the address of the management
interface.
Media failures on certain chunklets would sometimes prevent relocation
of other chunklets from the raid set as well as returning a preserved
LD back to normal. This has now been fixed.
Fixed chunklet relocation after media
failures.
60029
When a controller node rejoins cluster, HP 3PAR OS attempts to
rebalance the load across system. During the rebalancing process,
HP 3PAR system Panic - tpd:
Assertion point: file: cmp.c, line:
16883
60141
the current HP 3PAR OS does not handle the out of resource condition
which can cause entire system to be offline. The fix is to make sure
there is enough resource before starting the load rebalancing process.
There is a new database created in the “Atlantis" instance, called
"OAuth". Notice that there is a "space" in the name of the database.
Failed snapshots for atlantis cluster -
60272
For some reason, definitely a bug with the SQL writer, the writer fails
to come up; which explains why we don't see it in the list of vssadmin
writers. This topic has been discussed, and the user apparently found
the solution with no acknowledgement from Microsoft.
h
t
tp://
s
oc
i
al
.t
ec
hne
t
.mi
c
r
o
s
o
f
t
.co
m/F
o
r
ums/e
n
-U
S/dpms
qlbac
k
u
p/thr
ead/
4
3
2b4e3
8-6c7c
-
4f
e0
-a
be5-
7
6
9
2
7
5
7
8
4fcf/
If we have a heavily loaded system and have snapshots with lots of
data, the meta data tables which are heavily populated, then on
When deleting a very large and
filled snapshot can lock up a CPU for
long time
60494
deleting some of these snapshots the meta data tables are traversed
causing the cpu on occasions to be held for an excessive amount of
time. This can cause the locking out other processes/tasks from running
and can cause the background internode comms processing to be
delayed resulting in the erroneous belief that one of the other nodes
has died. This fix prevents the cpu from being held for an excessive
amount of time while traversing the the meta data tables during
snapshot removal. This fix includes an enhancement (65646) to cover
the case where the meta data tables are being merged during
snapshot removal. This fix with enhancement is applied to 2.3.1.MU5,
3.1.1.MU1, 3.1.2. The fix without the enhancement is in 3.1.1.(This
is also logged as # 52911).
Excessive raw space alerts have been seen when AO is configured
on a system. In 3.1.2 we have provided the ability to set raw space
alert threshold to 10g to suppress these alerts.
Raw space alerts
61058
It was possible to try to remove a VV that had been just removed from
an exported VV set before the VV export was fully cleaned up. In
some instances this would cause sysmgr to crash and restart.
Removal of volume from exported VV
set could cause sysmgr crash
61363
Modifications to the OS
9