EnterpriseSCHEDULE
Installation Notes
6.12 Version 7.1.16 corrections
- SCHEDULE V7.1.16 - 01
Problem: Variable expression case needed to translate consistantly.
Quotes not always needed. Solution: This has been fixed.
- SCHEDULE V7.1.16 - 02
Problem: Additional tracing around Memory Bounds warnings. Solution:
This has been implemented.
- SCHEDULE V7.1.16 - 03
Problem: Logging P queue "completions" "Purge Started" and "Purge
completed". Solution: This has been implemented.
- SCHEDULE V7.1.16 - 04
Problem: Incorrect Windows Gui mapping of Descripted node and IP node.
Solution: This has been fixed.
- SCHEDULE V7.1.16 - 05
Problem: Properly handling and corrections to Workgroup processing and
failover as directed by Customer Workgroup Beta Test Site. Solution:
This has been a mixture of fixes and implementation.
- SCHEDULE V7.1.16 - 06
Problem: Incorrect file handler logic on backlogged database
transactions. Solution: This has been fixed.
- SCHEDULE V7.1.16 - 07
Problem: Incorrect zero setid submissions. Solution: This has been
fixed.
- SCHEDULE V7.1.16 - 08
Problem: Allow the german character "ß" in schedule object names.
Solution: This has been implemented.
- SCHEDULE V7.1.16 - 09
Problem: Incorrects behavior in Windows GUI -resetproxy option.
Solution: This has been fixed.
- SCHEDULE V7.1.16 - 10
Problem: Change the GUI behaviour so that views opened when the last
GUI was closed are also opened when the -resetproxy option is used.
Solution: This has been implemented.
- SCHEDULE V7.1.16 - 11
Problem: Avoid asking twice for the password when -resetproxy is set on
the client and PROMPT_ALWAYS is set in the server parameter file.
Solution: This has been fixed.
- SCHEDULE V7.1.16 - 12
Problem: Fix the embedded command line client issues where password was
being asked. Solution: This has been fixed.
- SCHEDULE V7.1.16 - 13
Problem: Fix several workgroups problems in the GUI as directed by
Customer Workgroup Beta Test Site. Solution: This has been fixed.
- SCHEDULE V7.1.16 - 14
Problem: Added case insensitive node sorting in the Windows Gui monitor
view. Solution: This has been implemented.
- SCHEDULE V7.1.16 - 15
Problem: Change the progress default property page displayed for jobs
in the M and P queues.Switched from the statistics page to the state
page. Solution: This has been implemented.
- SCHEDULE V7.1.16 - 16
Problem: Problem arose when connecting to a secondary server which
didn't have a proxy set. Solution: This has been fixed.
- SCHEDULE V7.1.16 - 17
Problem: Fix to allow the entering of a negative value in the variable
field in resources, end actions, initiates and prerequisites dialogs.
Solution: This has been fixed.
- SCHEDULE V7.1.16 - 18
Problem: Change the default directory where the GUI log files are
created in case users don't have write access to schedule_library.
Solution: This has been implemented.
- SCHEDULE V7.1.16 - 19
Problem: Fix the creation of remote prerequisites in the layout and the
displaying of prereqs. Solution: This has been implemented.
- SCHEDULE V7.1.16 - 20
Problem: Fix relative path issue in connections. Solution: This has
been fixed.
6.13 Version 7.1.14 corrections
- SCHEDULE V7.1.14 - 01
Problem: Change to prototypes menu to have submenus for grouped jobs
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 02
Problem: Fix the comment field display in the job general property
page. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 03
Problem: Change the way the submit node of a created job is selected
using these guidelines.
- Use submit node of prototype job if created from prototype
- Use submit node of default job if in current directory
- Use work group of current directory if not empty and work group
active
- Use submit node of default job if present in parent folder
- Same as (3) above but for parent folder
- Use submit node of default job if present in default folder
- Use local node
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 04
Problem: Modified and corrected various issues for the display and
creation of workgroup replicated folders. Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 05
Problem: Properly handle empty submit node on OpenVMS Solution: This
has been fixed.
- SCHEDULE V7.1.14 - 06
Problem: Corrected rescheduling flags in the job property page.
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 07
Problem: Corrected 2 problems when there is a firewall between the
client and the server and the network names are not the same. Solution:
This has been fixed.
- SCHEDULE V7.1.14 - 08
Problem: Fix to better handle small/large fonts in the monitor view.
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 09
Problem: Corrects issue with the PID/Batch ID not being shown in the
GUI monitor for the queues K and L. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 10
Problem: Enhanced tracing in the Windows GUI. Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 11
Problem: Float problem when switching job to or from VMS Solution: This
has been fixed.
- SCHEDULE V7.1.14 - 12
Problem: Changed demo b scripts on all platforms to accommodate new
default prerequisite behavior. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 13
Problem: Error occurs while posting to queue case such that error
prevent future processing in schedule. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 14
Problem: Seg fault on Tru64 Unix when performing a TCPIP data send.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 15
Problem: Enhanced node not reachable retry logic. Solution: This has
been implemented.
- SCHEDULE V7.1.14 - 16
Problem: Enhanced TCPIP send message for VMS system. Better retry logic
and node not reachable logic. And logging for both reconnects (auto
logging), trace-bit setting for detailed logging, and additional server
dump logging. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 17
Problem: Multiple initializations of master control common once.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 18
Problem: Enhanced idle network disconnect logic. Keep idle Tcpip
connections longer and separate time stamps for decent and Tcpip
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 19
Problem: Enhanced channel disconnect logic by dumping channel details
into log file when abnormal disconnects happen. Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 20
Problem: Enhanced reconnect logic and added channel details to server
file when server reconnects happen. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 21
Problem: Corrected Link Control Block time stamp updates for better
logging and retry logic. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 22
Problem: Enhanced schedule_start.ksh/bash. Added signal to perform
server dump always on server restart because we were constantly missing
server details at customer sites. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 23
Problem: Corrected TCPIP retry connect/send anomaly for Windows and
UNIX. Previous code to segregate one request to not buildup broke code
that needed to continuously retry persistent requests. Solution: This
has been implemented.
- SCHEDULE V7.1.14 - 24 Problem Corrected backlog of request to
remove Qdj from Schedule Event queue when a job was deleted. Large
database deletes of jobs could adversely affect a schedule server. This
changed removed the excessive queue delete requests from a job delete.
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 25
Problem: Event log was showing only job name in job events, changed to
use full key value. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 26
Problem: Changed windows job completion status text from "Errorlevel
success"/"Errorlevel failure" to "Exited with a success code of
!UL"/"Exited with a failure code of !UL" Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 27
Problem: Command preprocessing discarded all commands when ((endif *))
was missing. This affected "type/notice/preprocess", "preprocess", and
job submit code when the job had preprocessing. Solution: This has been
fixed.
- SCHEDULE V7.1.14 - 28
Problem: Made the server flag "-r" group specific. Solution: This has
been implemented.
- SCHEDULE V7.1.14 - 29
Problem: Job create process code was broken. Logon or Logoff causes the
DACL on the windowstation and desktop to be reset. When this happened,
jobs would begin to fail. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 30
Problem: Add less ACE entries on the Windowstation DACL (from 2 per
user to 1 per user). Solution: This has been implemented.
- SCHEDULE V7.1.14 - 31
Problem: Added Window station and desktop DACL to "SCHEDULE SERVER
DUMP" command output. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 32
Problem: notify.c was not freeing the hProcess handle (leaking
resource). Solution: This has been fixed.
- SCHEDULE V7.1.14 - 33
Problem: Show group number in job log files (VMS,UNIX,Win32) Solution:
This has been implemented.
- SCHEDULE V7.1.14 - 34
Problem: Created schedule Report prototypes: prototype_reports.schedule
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 35
Problem: Created ntbackup prototype job Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 36
Problem: Sun stack trace now has function names (not just addresses)..
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 37
Problem: Notify errors were causing server to restart. Now they are
informational. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 38
Problem: Probe file failing when default device/directory/name was a
null_descriptor. Error was "Invalid string class". Solution: This has
been fixed.
- SCHEDULE V7.1.14 - 39
Problem: Journaling filenames shortened to guarantee they do not exceed
39 chars. Only a problem on VMS, but change made for all platforms out
of consistency. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 40
Problem: Need Job HOLD / NOHOLD functionality for Windows and UNIX
schedule servers. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 41
Problem: Schedule_notify.com: Text after BEGIN_/END_ block was ignored
on VMS. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 42
Problem: Schedule_notify.com, schedule_notify.vbs: MAIL_SUBJECT would
fail if the rhs string was quoted. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 43
Problem: Schedule_notify.com: control phrase lines would print to the
output stream. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 44
Problem: No-retry TCP send requests would begin retrying. Solution:
This has been fixed.
- SCHEDULE V7.1.14 - 45
Problem: The event monitor for a job was not showing events from nodes
in the same cluster. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 46
Problem: * Retry TCP send requests would sometimes do zero retries.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 47
Problem: History would report 'invalid data' when there were blank
lines in the file.. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 48
Problem: Need Job HOLD / NOHOLD functionality for Windows and UNIX
schedule servers. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 49
Problem: The event monitor for a job was not showing events from nodes
in the same cluster. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 50
Problem: Journal records were not recording the key index field in the
journal header portion. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 51
Problem: Connection issue between little endian/big endian machines in
rmt handlers. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 52
Problem: Access Violation in lclfileop.c; symptom in log is 'Dropped
nrq' type messages followed by access violation. Solution: This has
been fixed.
- SCHEDULE V7.1.14 - 53
Problem: Change workgroup activity code to be send-only. Reduces
problems with activity backlogging. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 54
Problem: Create/from was broken when the commands/notice text/notes
section had more than one SDL. (Unix/Win). Solution: This has been
fixed.
- SCHEDULE V7.1.14 - 55
Problem: Tie sys_caller in ssr_nrq_release to debug bit. Solution: This
has been implemented.
- SCHEDULE V7.1.14 - 56
Problem: 'Invalid ip channel' bug on VMS (server stops, does not
restart). Solution: This has been fixed.
- SCHEDULE V7.1.14 - 57
Problem: Exception caps around IP code on vms, prevents exceptions
under AST from stopping server. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 58
Problem: Fixed the command line client so that when it creates jobs, it
uses the os type of the server to determine how to set submit settings
such as submit flags, CLI, node, batch queue, etc. Problem was that if
client was on one platform & did set host to another platform and
created a job, the wrong settings were applied. Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 59
Problem: Workgroup wizard not working correctly. Solution: This has
been fixed.
- SCHEDULE V7.1.14 - 60
Problem: Modified translation of special characters. Used _5B instead
of _40 for the '@' character. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 61
Problem: Drag&drop and copy/paste problem. Solution: This has been
fixed.
- SCHEDULE V7.1.14 - 62
Problem: Changes to add "No access" as a choice to the ACL pulldown
menu. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 63
Problem: Allow scrolling horizontally in the job fields of the
connection property dialog in case the job name is too long to be
displayed entirely. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 64
Problem: Problem with variable values not shown in the layout when
using the explorer function "open as a layout" for a folder. Solution:
This has been implemented.
- SCHEDULE V7.1.14 - 65
Problem: Problem with window title cut off because the object name
& directory was too long. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 66
Problem: Changed default value for "Check Prerequisites" to unchecked
in the submit dialog Solution: This has been implemented.
- SCHEDULE V7.1.14 - 67
Problem: Error when modifying protections in the job access property
page. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 68
Problem: Improve an error message in the GUI to include the submit node
when a job submit fails Solution: This has been implemented.
- SCHEDULE V7.1.14 - 69
Problem: Speed up the drawing of large layout containing lots of jobs
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 70
Problem: New dialog "copy folder" used to to copy data from one folder
to a new folder, new option "New folder from" in the tree view under
nodes and level 1 folders Solution: This has been implemented.
- SCHEDULE V7.1.14 - 71
Problem: Add the display of the average time used per server call and
per node in the GUI log file when clicking on the "dump" button for
Nodeinfo. Performance flags can be selected from the debug window
(ctrl-shift F6). Solution: This has been implemented.
- SCHEDULE V7.1.14 - 72
Problem: Check in the notify script if the Messenger service is
running, necessary for the command "net send" to work. Solution: This
has been implemented.
- SCHEDULE V7.1.14 - 73
Problem: Earlier, the maximum size of text was limited to about 30K,
now it is limited to the maximum sdl size which is 256 * 1900 bytes
which is a little less than 500 K. This limit of 30K was only in the
mentioned property pages, it was already possible to edit larger text
in the edit window. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 74
Problem: Gui needs to handle the case when the replicate workgroup of
\syscontrol or \syscontrol\workgroup doesn't contain the node itself.
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 75
Problem: Increase the size of the fields "lead time for B entries" and
"Retention of P entries" in the monitor option. Solution: This has been
fixed.
- SCHEDULE V7.1.14 - 76
Problem: Change to reduce the QDJ delete requests when
deleting/renaming jobs. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 77
Problem: Add log file number handling in the GUI to show the log file
corresponding to the selected event in the monitor view. The field "log
file number" has been added to the progress window Solution: This has
been fixed.
- SCHEDULE V7.1.14 - 78
Problem: New monitor option to allow the display of centi seconds in
the monitor. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 79
Problem: Allow the job and event monitor views to group P and M
states/events by failure and success when sorting by state. Solution:
This has been implemented.
- SCHEDULE V7.1.14 - 80
Problem: Add the reconfigure value to the layouts. This new option
allow one layout edited manually to be reconfigured automatically when
loaded by the GUI. To reconfigure it horizontally use "Reconfigure=1",
for vertically "Reconfigure=2". If the layout is saved the
"reconfigure" value will be reset to 0 so it isn't reconfigured next
time it is loaded. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 81
Problem: Fix the server proxy reset command. Solution: This has been
fixed.
- SCHEDULE V7.1.14 - 82
Problem: Fix a GUI –reset proxy problem. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 83
Problem: Problem with the CLI in the submit property page. Solution:
This has been fixed.
- SCHEDULE V7.1.14 - 84
Problem: Check for the disable bit when reading a user record so that
disable users cannot connect anymore Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 85
Problem: Problem with columns being repeated in the explorer list view.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 86
Problem: Problem with the delete and rename animations, replaced by XP
animations. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 87
Problem: The event monitor for a job was not showing events from nodes
in the same cluster. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 88
Problem: Need Job HOLD / NOHOLD functionality for Windows and UNIX
schedule servers. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 89
Problem: Add log file number handling in the GUI to show the log file
corresponding to the selected event in the monitor view. The field "log
file number" has been added to the progress window Solution: This has
been fixed.
- SCHEDULE V7.1.14 - 90
Problem: Display the error message and not only the error code when
connection to a cluster member fails Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 91
Problem: Set the default value of "start initiates" to ON when submit a
job if the control is not displayed. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 92
Problem: Allow the GUI to open really large log files. Solution: This
has been implemented.
- SCHEDULE V7.1.14 - 93
Problem: Improve handling of loss connection to primary or secondary
servers. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 94
Problem: New icons and toolbars in the GUI. Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 95
Problem: New menu look. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 96
Problem: Modification of the variable property page and wizard to
handle expressions. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 97
Problem: New help menu option to show the job states help page from the
layout. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 98
Problem: Job log file and a&b scripts are created with the
ownership of the submit user (rather than job owner). Affects all UNIX
platforms. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 99
Problem: On tru64 UNIX a deadlock can occur when certain system level
utilities are invoked at the same time. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 100
Problem: Schedule server lacks ability to handle negative versions of
job log file. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 101
Problem: Need ability for UNIX platforms to include hundredths of
seconds in the date structure. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 102
Problem: Access violation on sun sparc in the type/que/full command:
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 103
Problem: General UNIX problem where elapse time is reported as zero in
the history record for all jobs Solution: This has been fixed.
- SCHEDULE V7.1.14 - 104
Problem: Log files being truncated by lines > 512 on VMS. Now, the
line will be discarded, but the rest of the log file will be read in.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 105
Problem: On going problem we have seen at customer site, where tcp/ip
communication stops between two machines (one machine can send to the
other, but no reply ever comes back). The guess is that we create the
socket for receiving, assign it to an lcb, and at the same time a
background thread determines that the lcb is inactive and releases it.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 106
Problem: CPU usage statistics for the Sun platforms. Solution: This has
been implemented.
- SCHEDULE V7.1.14 - 107
Problem: Journaling filenames that are > 39 chars, are a problem on
some VMS file systems. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 108
Problem: Need prototype jobs for standard Schedule reports. Solution:
This has been implemented.
- SCHEDULE V7.1.14 - 109 Problem Miss leading completion messages.
Previously they were "Errorlevel success" and "Errorlevel failure". Now
they will be "Exited with a success code of !UL" and "Exited with a
failure code of !UL". Solution: This has been fixed.
- SCHEDULE V7.1.14 - 110
Problem: Preprocessing problem when an ((endif *)) is missing.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 111
Problem: In creating a process on Windows the windows station logic
could be better managed like few DACL used and current D ACL be better
recycled. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 112
Problem: There is a need to add flexibility in formatting the
notification text. Solution: This has been implemented.
- SCHEDULE V7.1.14 - 113
Problem: Problem on Tru64 where "schedule whoami" (actually any
schedule command that tries to make use of sys_getjpiw()) in a batch
file fails. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 114
Problem: file access denied is incorrectly interpretted as an access
violation, causing a server restart Solution: This has been fixed.
- SCHEDULE V7.1.14 - 115
Problem: Need to make log files on shares viewable on Windows.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 116
Problem: Problem on all platforms where log files are left open on
certain cases. On VMS problem exposed because of an exhausted resource
error. Solution: This has been fixed.
- SCHEDULE V7.1.14 - 117
Problem: Need to increases a string size in a notification message. The
node name was being truncated in some cases Solution: This has been
implemented.
- SCHEDULE V7.1.14 - 118
Problem: Need to correct order of Schedule signal /complete such that
it occurs after the post processing script. Also, need parameter to
allow customer to revert back to old way where signal /complete occurs
before post processing in case it affects their way of running.
Solution: This has been fixed.
- SCHEDULE V7.1.14 - 119
Problem: Need to enhance Crontab conversion script to cover more cases.
Solution: This has been implemented.
- SCHEDULE V7.1.14 - 120
Problem: Need prototype job for application Banner. Solution: This has
been implemented.
- SCHEDULE V7.1.14 - 121
Problem: potential access violation that can occur at the below
location: