/temp filesystem is full

I apologise for being a plonker, but I have a problem installing Solaris on a Compaq E500 laptop.
I try to install Solaris 8 X86 version on this laptop following Sun's instructions to the letter until it gets to the FDISK stage then problems occur.
The laptop has an 18GB hard disk and the BIOS does not correctly report the size of the hard disk to FDISK. FDISK says that it is only 8GB big. I have flashed the BIOS with the latest version from Compaq but it still has the same results.
However, FDISK does let me create Solaris partitions and continue the installation. Everything is fine until we get to the section "Copying mini-root", preparing to eject media and restart etc, then the message /temp filesystem is full (something like that) appears several times and the machine then locks. This happens regardless of how big I set the swap partition.
I am making a big assumption here that this is caused by the hard disk size not being correctly reported. Is there anyway around this problem?
Many thanks,
Richard.

Richard,
Two things: First, try booting the "Software 1 of 2" cdrom rather than the "Installation" one. This skips the whole "copying miniroot" business and runs from the cdrom miniroot during the install.
Second; make sure you are using the latest possible DCA floppy if you are using one.
There may still be driver issues you need to address on the E500 itself but these two step should get you going in the right direction.
Don

Similar Messages

TEMP tablespace getting full while inserting a CLOB in Trigger

We have a Oracle 10g (10.2.0.4.0) DB on a Solaris 9 box which also runs our J2EE web-service application on Weblogic 8sp6 server.
We get around 220K web-service requests from upstream callers daily to insert data in the main table, say TABLE1, which has daily partitions on a date column. This table has around 21 columns out of which 1 is a CLOB column.
Now this table has an AFTER INSERT trigger which calls a package procedure to insert the same record into another table, say TABLE2.
From Java application insert statement in executed in below format using a weblogic jdbc connection pool :
INSERT INTO TABLE1(COLUMN1, COLUMN2, ........., CLOB_COLUMN,........, COLUMN21) VALUES (:1, :2, :3, :4, :5, :6, :7, :8, :9, :10, :11, :12, :13, :14, :15, :16, :17, :18, :19, :20);
Clob object is prepared in application using ojdbc14.jar.
We are observing a strange issue here. The TEMP tablespace utilization keeps on growing as more and more inserts are executed by application and after ~125K inserts the TEMP tablespace gets full and we start getting ORA-01652 error.
On further analysis we could see that there are only 7-10 session being maintained but as more and more inserts happen TEMP tablespace utilization goes on increasing for each of these sessions.
When we tried with inserting just few records and then watching the session details in v$session_wait then we could see that it is in INACTIVE state and waiting for the event ‘SQL*Net message from client’. This does not seem correct as the session has successfully inserted the data and committed the transaction and we can see the data in the tables as well.
The confusing thing here is when we modify the trigger to pass blank string('' ) instead of the CLOB column to TABLE2 then this issue does not occur. All 200K records are inserted properly and TEMP tablespace utilization also keep always below 1%.
Can you please help us in solving this issue. Is this related to any oracle issue?
Inside the package we have tried using DBMS_COPY statement to copy the CLOB column after insert but still same result.
Code for reference:
Trigger:
=====================================
CREATE OR REPLACE TRIGGER trg
AFTER INSERT OR UPDATE
ON TABLE1
REFERENCING NEW AS NEW OLD AS OLD
FOR EACH ROW
BEGIN
IF (:NEW.date_col > SYSDATE - 2)
THEN
IF (:NEW.cat IN (1001, 1002))
THEN
pkg.process_change
     (:NEW.COLUMN1,
          :NEW.COLUMN2,
          :NEW.CLOB_COLUMN,
FLAG
END IF;
END IF;
END;
=====================================
Package:
=====================================
procedure PKG.Process_change(
p_COLUMN1 number,
p_COLUMN2 varchar2,
p_CLOB_COLUMN clob,
flag boolean
) is
v_watermark pls_integer;
v_type varchar2(1);
begin
if (flag) then
v_type := 'U';
else
v_type := 'I';
end if;
select t_seq.nextval into v_watermark from dual;
insert into TABLE2(
COLUMN1 number,
COLUMN2 varchar2,
CLOB_COLUMN clob,
watermark,
dml_type
)values (
p_COLUMN1 number,
p_COLUMN2 varchar2,
p_CLOB_COLUMN clob,
v_watermark,
v_dml_type
end;
=====================================

My first thought on reading your post is that not only are you using a database version that is now so old it is in extended support and even then not even the most recent patchset for it.
The first thing I would do is move to 11gR2 and if you can't do that at least get to 10.2.0.5 and apply CLOB relevant patches as well.
Same goes for your operating system. Solaris 9 is ancient: So move to 10 which has vastly improved memory management.
To help you further it would be really valuable to know the table layout. For example is this a heap table or an IOT? Is it partitioned? Is this RAC? What size are the CLOBs? Are they stored in-line? Chunk size? etc.
This page should start you down the right road:
http://docs.oracle.com/cd/B19306_01/appdev.102/b14249/adlob_tables.htm#sthref204
But I am also wondering why you would use a trigger to, as you say, "insert the same record into another table." This description is a poster child for "bad design."

/var filesystem 100% full

I need some help here please, I cleared /var filesystem from all the big files but it's still showing me 100% full, I only manage to get rid of this problem when I reboot the server then /var becomes 3% full, any idea what's causing this to happen.
I want to avoid rebooting the server in future, should /var/ become full again
Many thanks

Some process still has the deleted file open, so the kernel cannot
free the space occupied by the file(s) you've delete.
The space is deleted as soon as the process closes the file handle
(e.g. by stopping / restarting the process).
The following command lists the process ids / file handles that
do not have filesystem links but are still referenced by a process
(you must run it as user root):
# find /proc/*/fd -type f -links 0 -ls

Temp tablespace 99%full

What are the suggestion for temp tablespace, if its 99%full, should we keep adding space in it or it will use existing space?
TABLESPACE TOTAL_MB USED_MB FREE_MB PCT_USED GRAPH (X=5%)
TEMP 17,500.00 17,444.00 56.00 99.68 [XXXXXXXXXXXXXXXXXXX-]

Hi,
>>What are the suggestion for temp tablespace, if its 99%full, should we keep adding space in it or it will use existing space?
I think that you don't need worry about, unless you're receiving some ORA- error about out of space in TEMP tablespace. You can see below an SQL output from a database used here in my company for development and tests purposes. The database is up uninterruptedly by 7 months and the space size for the TEMP tablespace have been configured to use 900 MB.
LEGATTI@ORACLE10> SELECT tablespace_name, SUM(bytes_used), SUM(bytes_free)
2 FROM   V$temp_space_header
3 GROUP BY tablespace_name;
TABLESPACE_NAME                SUM(BYTES_USED) SUM(BYTES_FREE)
TEMP                                 943718400               0Cheers
Legatti

Temp tablespace 98% full

Hy
My temp tablespace (1 GB) is 98% full.
A.)How can I find out which user uses at a certain point of time in percent the tablespace.
B.)Does a full temp tablespace lead to performance problems reduces the "reply velocity" of a database
lutz
null

Lutz,
a) By using the following query you can determine who are all the users who have created their objects in TEMP tablespace..
SQL> SELECT OWNER, SEGMENT_NAME, SEGMENT_TYPE
from dba_segments
where tablespace_name like 'TEMP';
Once u get the names of the objects u can move to the tablespace u want.
b). According to me YES. Try to restart the database after moving the objects if any.. and this should give some space to u.
There is a bug in 8.1.6 related to the use of TEMP tablespace while using the sort operations.. Check it if u r using 8.1.6
Hope this helps,
Shiva
null

Temp db is full

Hi all. Please let me know what needs to be done in below scenario.
TempDB is full. Unable to expand the databases and unable to open a new
query window. We should not restart sql services or restart the server.
How to fix this.

Hi all. Please let me know what needs to be done in below scenario.
TempDB is full. Unable to expand the databases and unable to open a new
query window. We should not restart sql services or restart the server.
How to fix this.
First of all that is not a very good question in sense that if you would have searched a bit you could have found a answer specially How to
Troubleshoot Insufficient disk space on Tempdb
You have to anyhow add additional disk space to tempdb data file or
add an extra file on drive where space is present so that tempdb can grow and transaction which requires tempdb can complete than you can decide further course of action. Adding a disk space or additional drive is correct solution here.
You should place disk monitoring in place to inform you when disk space crosses particular threshold you should also find why transaction log file of tempdb grew out of proportion
Please mark this reply as answer if it solved your issue or vote as helpful if it helped so that other forum members can benefit from it
My Technet Wiki Article
MVP

Root filesystem reporting full [SOLVED]

Can anyone explain why this is?
df
Filesystem 1K-blocks Used Available Use% Mounted on
udev 10240 0 10240 0% /dev
run 10240 596 9644 6% /run
/dev/disk/by-uuid/041fef4d-ebc5-4e25-b38c-2f6a0fafafef 7566432 7419700 0 100% /
shm 958956 1600 957356 1% /dev/shm
tmpfs 958956 12 958944 1% /tmp
/dev/sda4 232433560 68292028 152334564 31% /home
/dev/sda1 101086 18635 77232 20% /boot
Last edited by dejakaymac (2011-10-20 08:28:40)

karol wrote:
dejakaymac wrote:Can anyone explain why this is?
Sure: you have too much stuff there. If you''ve installed a ton of packages maybe you need a bigger / partition.
Check for some rogue logs growing to a couple GB.
Or, if you have been doing a lot of experimenting with different apps, ie. installing and removing to find what you like, your package cache might be quite large.
pacman -Sc will clean it up for you.
Cheers.

Df not updating --filesystem says its full

Hi everyone. I have the strangest problem. du says my home directory contains less than 100M. df says its full (1GB). Even though df says its full, i can still create about 50M worth of files there. Then when i delete the files, df still says the filesystem is full. Any help would be very much appreciated. This is OEL5 and the last thing i did was install sqldeveloper.
==========================================================
[root@ajax home]# cd /home
[root@ajax home]# du -h --max-depth=1
32K ./private
748K ./private
16K ./lost+found
36K ./private
13M ./oracle
32K ./private
14M .
[root@ajax home]# df -h |grep home
/dev/mapper/vg00-lvhome 992M 946M 0 100% /home
[root@ajax oracle]# cat TEST>>TEST2
[root@ajax oracle]# cat TEST2>>TEST
[root@ajax oracle]# cat TEST>>TEST2
[root@ajax oracle]# cat TEST2>>TEST
[root@ajax oracle]# cat TEST>>TEST2
cat: write error: No space left on device
[root@ajax home]# ls -lh TEST*
-rw-r--r-- 1 root dba 21M Aug 19 08:06 TEST
-rw-r--r-- 1 root dba 26M Aug 19 08:06 TEST2
[root@ajax oracle]# rm TEST*
rm: remove regular file `TEST'? y
rm: remove regular file `TEST2'? y
[root@ajax home]# df -h |grep home
/dev/mapper/vg00-lvhome 992M 946M 0 100% /home
So neither du nor df seems correct. Df clearly isn’t correct because I can create files and then after I delete them it still reports the same thing. Du doesn’t seem correct or I should be able to create about another 900M of files on /home

Hi DaveLehr,
One of the possibility is the file you removed for space was being used by some process/application.
Linux will allow you remove that without any warning.
After that, the file is invisible by "ls -l", or "du -h --max-depth=1"
but the space of that file won't be released until the last process who is using that file close it.
After that, you can see as you expect from "df -h"
since you've already rebooted ur machine, This is only one of the possibility. :)
see as below:
# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
9.5G 8.6G 377M 96% /
# dd if=/dev/zero of=zero bs=1M count=300
# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
9.5G 8.9G 77M 100% /
# du -h --max-depth=1
301M .
# open another console and edit the file, don't close it
# vi zero
# lsof | grep "tftpboot/zero"
vi 15534 root 3r REG 253,0 314572800 196618 /tftpboot/zero
# rm -f zero
# lsof | grep "tftpboot/zero"
vi 15076 root 3r REG 253,0 314572800 196618 /tftpboot/zero (deleted)
# du -h --max-depth=1
76K .
# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
9.5G 8.9G 77M 100% /
# switch to the other console and close it
# exit vi
# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
9.5G 8.6G 377M 96% /
Message was edited by:
rwen

Temp segments full

Hi,
There is a report based on a select which includes several tables, views and an inner select. In the last several years, any date will give no more than a couple of hundred records except for one day in 2006 when there are almost 3,000 records, and November 8th of this year when there are over 7,000 records.
The report is fine except when it includes those two days. However, running the select in plsql is no problem at all (results within a couple of seconds). It is only in Discoverer where a message will come up that the temp segments are full.
Can anyone give me a suggestion as to what to look for in order that I can change the sql?
Thanks.
Leah

Hi Srini,
Thanks for responding. In the meantime I received a response from Metalink. Apparently the sql that is being formed by discoverer is different than the custom sql folder. I am supposed to run the sql that was executed in Discoverer in sql*plus. I am waiting for sql*plus to be installed in my computer at work.
Thanks. I will be happy to update this thread with the results.
Leah

Temp tablespace full at startup

Hi,
my temp tablespace appears full at database startup. I know it is full because the collecting statistics process can not be executed since temp tablespace can not grow (autoextend off). I had to add another temp datafile to get the statistics but first one continues full after each restart.
Is this any signal of malfunction?
Thanks for all.
Greetings.

ok. I think all is understood.
DBMS_STATS.GATHER_SCHEMA_STATS probably needed more disk than 1.5 GB (my old temp tablespace), so the result was
ERROR en línea 1:
ORA-01652: no se ha podido ampliar el segmento temporal con 128 en el
tablespace TEMP
ORA-06512: en "SYS.DBMS_STATS", línea 9136
ORA-06512: en "SYS.DBMS_STATS", línea 9616
ORA-06512: en "SYS.DBMS_STATS", línea 9800
ORA-06512: en "SYS.DBMS_STATS", línea 9854
ORA-06512: en "SYS.DBMS_STATS", línea 9831
ORA-06512: en línea 1
Then adding new datafile (1.5 GB more) process could finish. Database continues growing, process always had finished not yesterday: the day of needing full temp tablespace came.
Tell me if I am wrong.
Thanks a lot.

[Cisco FAQ] - Full filesystem

One of my DNCS filesystems is full. What is the best way to start cleaning it up?

One suggeston would be to run the following command from the mount point of the filesystem to be cleaned up:
SP Video TAC Lab SR 4.2>> du -h | sort -rk 1,1 | grep -v "K" | less
929M .
9.1M ./spectrum/SS-Tools/dbpart1
9.1M ./dncs/downloads/SAIgqam/reloc
9.1M ./dncs/downloads/SAIgqam
Then move into the larger directories and run the above command again until large files that can be deleted are found. Repeat in other large directories, as needed.
Alternatively, the command below can be run repeatedly to idnetify where the filesystem space is being used:
SP Video TAC Lab SR 4.2>> du -s * | sort -n
0 lost+found
10 jdbc
14 easftp

Application publishing issues after temp profile usage

We haven't figured out the solusion to this issue but we do know the cause and a workaround for it that works, kind of...
A brief description would be that after permission issues for some users roaming profile storage on the fileserver application publishing and several applications fail in a different way for different users. This is due to the creation of C:\Users\TEMP,
TEMP.DOMAIN etc folders rather than using C:\Users\Username folders. The issue seems to be that App-V 5 client remembers this temporary folder in a way that we havent figured out yet, all references to the temp folders for each users account UID in the
registry has been removed and the roaming profie was deleted locally and on the fileserver YET app-v remembers the path C:\Users\TEMP and tried to create shortcuts etc there without success --> the packag then failes completely and dissapears from the client/powershell
Get-AppvClientPackage view as if its not even published to the user.
The BIG question is how the "#¤/&¤ can you get the App-V client to forget about the C:\Users\TEMP foders and retun to the users correct folder? We have tried to remove all user accounts UID from HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\AppV\MAV\Configuration\Packages
etc and also the users profiles so this data must be stored somewhere else. The issue is local to specific XenApp servers, if the temporary profile wasnt created on a server the issue will never appear there.
The workaround is that all users with the issue have full access to the paths C:\Users\TEMP and C:\Users\TEMP.DOMAIN folders, however when app-v generates shortcuts in that folders we'll need to manually move them to the correct path so that they are
available to the users, smooth!
The environment is based on 4x Windows 2008 R2 server with citrix XenApp 6.5. App-V 5 SP1 RDS client is installed and configured to use a single publishing server "Add-AppvPublishingServer -Name SERVERNAME -URL
http://FQDN:8080"
The details:
(we think) Some users had the follow error since they couldn't access the faulty profile path TEMP or TEMP.DOMAIN:
Failed to create file by callback data: \Device\HarddiskVolume2\Users\USERNAME\AppData\Roaming\Microsoft\AppV\Client\VFS\FE7FE758-FDD8-4208-AF3D-E2505AEF6791\AppData\Adobe\Acrobat\
Another user got only one package (the only package without a shortcut):
This user receives 1 out of 8 packages.
We have tried removing the local and centrally stored roaming profile without any change.
The following is recorded in the event viewer (Microsoft\AppV\Client\Admin):
Log Name:      Microsoft-AppV-Client/Admin
Source:        Microsoft-AppV-Client
Date:          2014-02-04 09:30:04
Event ID:      19104
Task Category: Publishing Refresh
Level:         Error
Keywords:      Publishing Refresh
User:          DOMAIN\username
Computer:      XENAPPSERVER1.domain.com
Description:
Part or all packages publish failed.
published: 1
failed: 8
Please check the error events of 'Configure/Publish Package' before this message for the details of the failure.
Event Xml:
<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
<System>
    <Provider Name="Microsoft-AppV-Client" Guid="{E4F68870-5AE8-4E5B-9CE7-CA9ED75B0245}" />
    <EventID>19104</EventID>
    <Version>1</Version>
    <Level>2</Level>
    <Task>36</Task>
    <Opcode>0</Opcode>
    <Keywords>0x8000000000800000</Keywords>
    <TimeCreated SystemTime="2014-02-04T08:30:04.164025400Z" />
    <EventRecordID>661</EventRecordID>
    <Correlation />
    <Execution ProcessID="3376" ThreadID="34620" />
    <Channel>Microsoft-AppV-Client/Admin</Channel>
    <Computer>XENAPPSERVER1.domain.com</Computer>
    <Security UserID="S-1-5-21-XXXXXXXX" />
</System>
<EventData>
    <Data Name="Value1">1</Data>
    <Data Name="Value2">8</Data>
</EventData>
</Event>
In addition to the above the following is recorded for every failed package (Microsoft\AppV\Subsystems-RegistryStaging), however the package that works arent displayed in the log.
Log Name:      Microsoft-AppV-Subsystems-RegistryStaging/Debug
Source:        Microsoft-AppV-Subsystems-RegistryStaging
Date:          2014-02-04 09:30:02
Event ID:      2002
Task Category: None
Level:         Information
Keywords:
User:          DOMAIN\username
Computer:      XENAPPSERVER1.domain.com
Description:
Configure package called on already configured package '{a849ce4c-2b7b-46a9-b54e-c72e61858de1}' version '{e780bf5c-effa-4419-bd4a-03bcc394baf4}'.
Event Xml:
<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
<System>
    <Provider Name="Microsoft-AppV-Subsystems-RegistryStaging" Guid="{1f4a896a-fea6-4862-bbeb-aaa190ae31c0}" />
    <EventID>2002</EventID>
    <Version>0</Version>
    <Level>4</Level>
    <Task>0</Task>
    <Opcode>0</Opcode>
    <Keywords>0x8000000000000000</Keywords>
    <TimeCreated SystemTime="2014-02-04T08:30:02.853894400Z" />
    <EventRecordID>89</EventRecordID>
    <Correlation ActivityID="{01AFCC48-F800-0001-4FF3-2C81241CCF01}" />
    <Execution ProcessID="3376" ThreadID="34620" ProcessorID="3" KernelTime="544" UserTime="3885" />
    <Channel>Microsoft-AppV-Subsystems-RegistryStaging/Debug</Channel>
    <Computer>XENAPPSERVER1.domain.com</Computer>
    <Security UserID="S-1-5-21-XXXXXXXX" />
</System>
<EventData>
    <Data Name="package">{A849CE4C-2B7B-46A9-B54E-C72E61858DE1}</Data>
    <Data Name="version">{E780BF5C-EFFA-4419-BD4A-03BCC394BAF4}</Data>
</EventData>
</Event>
We also notice the following errors in Microsoft\AppV\Subsystems-shortcuts when AppV tries to generate icons in the wrong/old profile path (cmd.exe "set" displayes the correct path but why cant appv get this info?):
Log Name:      Dbg
Source:        Microsoft-AppV-Subsystems-shortcuts
Date:          2014-01-28 13:55:47
Event ID:      130
Task Category: None
Level:         Error
Keywords:
User:          DOMAIN\username
Computer:      XENAPPSERVER1.domain.com
Description:
Failed to write shortcut file 'C:\Users\TEMP.DOMAIN\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\' with error '6629316385409335299'.
Event Xml:
<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
<System>
    <Provider Name="Microsoft-AppV-Subsystems-shortcuts" Guid="{a7782ffe-c5f5-4554-ab1c-acb874c1ccf9}" />
    <EventID>130</EventID>
    <Version>1</Version>
    <Level>2</Level>
    <Task>0</Task>
    <Opcode>0</Opcode>
    <Keywords>0x8000000000000000</Keywords>
    <TimeCreated SystemTime="2014-01-28T12:55:47.925264700Z" />
    <EventRecordID>251</EventRecordID>
    <Correlation />
    <Execution ProcessID="3376" ThreadID="3628" ProcessorID="0" KernelTime="378" UserTime="136" />
    <Channel>Dbg</Channel>
    <Computer>XENAPPSERVER1.domain.com</Computer>
    <Security UserID="S-1-5-21-XXXXXXXXXXXX" />
</System>
<EventData>
    <Data Name="string">C:\Users\TEMP.DOMAIN\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\</Data>
    <Data Name="uint64">6629316385409335299</Data>
</EventData>
</Event>
We noticed thet the user DOMAIN\user had a TEMP.DOMAIN profile path due to faulty permissions on the profile store and that this path thoug its fixed was stil configured in the AppV part registry according to below.
Since this key is global to the server and not bound to the user account it wasnt removed with the user profile, we tried removing the specific keys manually where the user account UID occured without any change to the problem. The removed keys werent
regenerated in the registry though.
The specific user account SID uses TEMP.DOMAIN paths (\Users\TEMP.DOMAIN\AppData\Roaming\Microsoft\AppV\Client\VFS) for two String Value names "TargetLongName" and "TargetShortNamefolders" in the following key (and others):
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\AppV\MAV\Configuration\Packages\04ADC3E8-E368-4B95-98AF-F53A625070CD_C248B175-B174-41F5-A4D6-62835CF86519\UserConfig\S-1-5-21-XXXXXXXXXX
As mentioned the UID keys "S-1-5-21-XXXXXXXXXX" did not reappear UNTILL we manually created the folder C:\Users\TEMP.DOMAIN with full permissions for the affected user. Once this folder was created App-V recreated the registry keys "S-1-5-21-XXXXXXXXXX"
with the correct \Users\USERNAME\* values however the icons for the applications was distributed to the C:\TEMP.DOMAIN\Desktop folder and since the users desktop is accually in C:\Users\USERNAME\Desktop none of the icons was visible to the user.
Another user had similar issues but with the java plugin for internet explorer failing without error messages, once that user got access to C:\Users\TEMP.DOMAIN the user was again able to run java.

Cheers,
Ended up putting in some workarounds.
Citrix\Microsoft best practice hotfixes
Review all userprofiles to check correct permissions etc (powershell script)
Purge Key each week, possibly switch to nightly.
$CitrixOU
=Get-ADComputer
-filter *
-searchBase 'OU=Citrix,OU=Servers,DC=test'
| Select-Object
-expand Name
Invoke-Command
-cn $CitrixOU {gci
HKLM:\SOFTWARE\Microsoft\AppV | foreach-object {gci
$_.pspath}
|where-object {$_.name
-like "*S-1*"}
|remove-item
-Recurse -Force}
until we can work out if this is 'by design'

IOH Temp - Obsession, Over Reaction, Or A Real Reason For Concern

Many of us have concerns over what is an acceptable IOH temp at idle & full load. Please, no rants, tangents, or bashing. If you feel your temp is too high, lets keep everything constructive, & let's all work together to get it fixed. To date, we have two types of boards to consider, those with push pins, & the others with machine screws. A simple test to see if your IOH temp may benefit from some form of a mod, a gentle but firm downward pressure on the IOH heatsink after it is fully warmed up, it may take 15 seconds or so before any reaction, but does the temp start to decrease? By how much? If around 10c reduction, increasing the downward pressure by hardware modification may be an option.
Before taking this too far, what 'is' too hot? If we can believe the manufacturer of the chip, 100c is it's max. limit. For everyday use at full sustained load, from what I can gather from Intel, 95c Tcontrol appears to be the 'magic' upper limit without any possible long term degradation or data corruption. Anything at or below that, as per them (meaning Intel), should have no effect on the chip's lifespan. The jury is still out, & only time will tell. Details can be obtained by reading the Intel X58 Express Chipset Thermal and Mechanical Design Guide of March 2009.
For those that wish to take further steps to reduce their IOH temps, these existing threads have a 'lot' of useful information on how to, what with, & where it can be found.
References; (note Reply #38 by mighty in Scotslad's thread) Scotslad https://forum-en.msi.com/index.php?topic=124752.0
evr999 https://forum-en.msi.com/index.php?topic=126885.0
Dankirk2 https://forum-en.msi.com/index.php?topic=129339.0
Also, voice your concerns if you feel it necessary to MSI Tech. Support. Depending on your location, USA & UK http://ocss.msi.com/ other areas of the world http://support.msi.com/ Keep in mind, MSI is 'very' aware of our concerns, & they are making changes. I've seen them on a couple of recent boards I have assembled.

Funny you should ask... At about the same time that the Xserve was discontinued, we also lost our channel rep of nearly 14 years. Right now we are not sure who our rep is. I have however, sent this to as many people in Apple that I could think of. And I plan on continuing to reach out to as many as possible.
The sad reality is that this decision is impacting customer deployments and the future of customer installations. Apple strives for market share and loves to tout those numbers. But the numbers we fought for over these last eight years are going to disappear.
As much as this impacts our business, we are making up for it in enterprise deployment of systems. We will survive and lead in our geographic area. I do not fear that. But OS X Server is a UNIX server and with it has come all of the benefits of a UNIX system, including legitimacy in a data center. The Intel Xserve cemented this reality because of its hardware configuration and form factor. Yes, we have G4 and G5s still in service. But those are mostly in all Mac shops that understand the benefits and can work around the buzz word limitations. But fortune 100/500 IT departments did not really get on board until the release of the Intel server.
Still beating the drum to raise awareness. Spread the word and talk to your Apple reps.
Also, ask about the direction that Apple's Enterprise team is going. If the news I hear is correct, then our fears about all things pro and enterprise going away are likely true.

How to determine what's using data store temp space?

How can one determine what's using data store temp space? We are interested to know what structures are occupying space in temp space and if possible what pid/process connected to TimesTen created them.
Also, is there a procedure that will work if temp space is full?
Recently one of our data stores ran of space. We we're unable to run commands like "monitor", "select * from monitor", "select count(*) from my_application_table", etc. These commands failed because they required temp space to run and temp space was full. We killed the application processes, this in turned freed up temp space, then we were able to run these queries.
Ideally, we'd like to have a procedure to figure out what's using temp space when temp space is full.
The other thing we could do is periodically monitor temp space prior to it filling to determine what's using temp space.

That was my original thought, but once you click the slider track or thumb, and then enter a value in the text control, the clickTarget on the change event envoked by the change to the bound data (after entering a value in the text control) will be whatever slider element had last been clicked. If you've never clicked the slider, clickTarget=null. But once you've clicked the slider the clickTarget always has a value of "thumb" or "track", regardless of what triggered the change event.

Can i prevend the generation of files in temp directory

I noticed, that a labview (7.0) application generates different files in the temp directory when executed and deleting them when stoped.
(Except of the *cur.txt and *log.txt files which only exists once per application)
Is there a way to prevend LabView to write this files (lvtl*.rsc; lvtr*.rsc; *.tmp)
I've got an application, which i have to kill (there's no other way at this moment).
If i do so, the files still exist. I kill this application about once a minute.
Because i don't end the application the regular way, the mentioned files still exist.
As a result the temp directory runs full.
Suggestions are appreciated.
Best regards,
Rainer
P.S.: The Suggestion to regular stop the application does not help me of course ;-)

Rainer,
Nice question!
There are different ways for doing this. Unfortunately no any "100% guaranteed" way (because its not documented by NI - which temporary files are created by LV app):
So, first suggestion with deletion of all temporary files seems to be safety, because normally Temp directory must be empty. In case if some files currently used by other applications, they locked and you cannot delete them.
Second way - change path to temp directory especially for your application. Just add tmpdir key into assotiated ini file:
[app]
tmpdir=C:\MyApp\Temp
After that lvtl*.rsc; lvtr*.rsc will be created into this directory. You can safety delete these files after kill.
(Note, that "*" is pid of your application)
Third way - looking for all handles, which currently opened by your application. You can do this with some utilities, for example with Handle.exe from SysInternals.
For example for my simple application following output was generated (some not important lines are skipped):
Handle v2.2
App.exe pid: 656
cc: File C:\Temp\App_7.0_Administrator_cur.txt
144: File C:\Programme\National Instruments\Shared\LabVIEW Run-Time\7.0\lvapp.rsc
148: File C:\MyApp\App.exe
1f0: File C:\Temp\nix2E5.tmp
210: File C:\MyApp\Temp\lvtr656.rsc
214: File C:\MyApp\Temp\lvtl656.rsc
760: File C:\MyApp
Now you can parse output and looking for all currently opened temporary files, then kill your your application and then safety delete tmp and rsc files. Note, that files lvtl*.rsc and lvtr*.rsc are located in "local" temporary directory as described in ini file, but nix*.tmp still in "default" directory.
Hope it help a little bit...
Andre.

/temp filesystem is full

Similar Messages

Maybe you are looking for