Quantcast
Channel: SCN : Discussion List - SAP Adaptive Server Enterprise (SAP ASE) for Custom Applications
Viewing all 685 articles
Browse latest View live
↧

(name NAME NOT RECOVERABLE) and MEMC_CRCONSTR_1

November 24, 2015, 3:45 am
$
0
0

We occasionally get the message

     Cannot run query--referenced object (name NAME NOT RECOVERABLE) dropped during query optimization.

 

This

Updated Writeup for ASE Error Message 225

 

suggests either number of open objects which looks ok

 

Name                            Num_free    Num_active  Pct_act Max_Used    Reuse_cnt   Instance_Name

----------------------------   --------------   --------------- ---------- --------------   --------------- ------------------------------

number of open objects          32035       17965    35.93        24629                       0  NULL

 

or Using a mix of qualified

 

I can't see any mixed qualifiers and its not easy to reproduce - seems to happen when the server is under heavy load.

 

Looking at the procedure cache I've noticed something odd with this entry in monProcedureCacheMemoryUsage

 

InstanceID AllocatorID ModuleID    Active      HWM         ChunkHWM    NumReuseCaused AllocatorName

---------- ----------- ----------- ----------- ----------- ----------- -------------- ------------------------------

          0          25           2           0  2147483646           2              0 MEMC_CRCONSTR_1

 

Could the 2 things be related ?

What is the MEMC_CRCONSTR_1  ( Sounds like creating a constraint).

↧
Search

How to capture database auto expansion event that is triggered by Sybase ASE ?

November 22, 2015, 7:53 pm
$
0
0

Hi All,

 

I'm back , I was very glad when retrieving a lot of extremely useful information in my previous post (How do I get the spid of Sybase when using JDBC). I did apply for my project and it has worked perfectly, thank so much for your supports. Today, I come back with some questions about Sybase device and database auto expansion functionality. Hopefully we will have an interesting discussion on this, and then we can find a good solution for my situation. Here is the questions:

 

1. Can we remove a device without deleting database first ? As far as I know, a device ( physical file on Operating System level ) is the first thing that need to be created before we do another things, such: creating segment, partition, create/alter database ....  Over the time, the data in database will grow more and more bigger, and we need to create a new device to store the data. But we also want to remove old device because it no longer uses, the only way to remove the old device is to delete database first and then removing the device. That is impossible for my product, I did google the solution and there is a suggestion about removing record in sysdevices and sysdatabase tables, but I'm not sure that will be good approach. Would you please share me your idea about this ?

 

2. Can we capture the database auto expansion event that is triggered by Sybase ASE ? I'm using the Database Auto Expansion functionality, that is provided by Sybase ASE ($SYBASE_HOME/ASE-15_0/scripts/installdbextend). When enabling this function, we can set the threshold, max device size and growth rate for a specific device. And then Sybase ASE will trigger automatically if threshold is reached, the only way to know the event has occurred or not via Sybase log file  ($SYBASE_HOME/ASE-15_0/install/<Database_Name>.log). I need to know when the event occurs for notification purpose, if you have any ideas on this, please share with me. I really appreciate !!!

 

Thanks

Khoa Tran

↧
↧

SYBASE ASE and IQ tuning question

November 26, 2015, 12:51 am
$
0
0

Hello All

 

We use an reporting solution which uses SYBASE ASE for management data and SYBASE IQ for reporting data.

While launching the reports we are finding the reports are loading very slow when multiple users are connected.

 

We have 8 CPU cores and 48 GB Physical memory running on RHEL Linux 6.x on VM environment.

 

SYBASE IQ database file system is mounted on SAN drive and is been accessed as cluster file system.

 

Sybase IQ db version is,

 

Sybase IQ/15.4.0.3046/141204/P/ESD 5/Enterprise Linux64 - x86_64 - 2.6.18-194.el5/64bit/2014-12-04 17:24:15

 

We checked and found that number of user connections supported are not more than 50 by sybase IQ, so we installed
an IQ patch which will increase the number of user connections.

 

Even after that we saw slight increase in performance, but it did not fit our requirements.

 

Attached SYBASE.cfg file and our IQ DB startup parameters are,

 

-gc 20

-gd all

-gl all

-cl 256m -ch 1024m

-gm 100

-gp 4096

-ti 10

-iqgovern 20

-gn 100

-iqnumbercpus 8

-iqmt 1024

-iqtraceflags 2

-iqmc 26624

-iqtc 13312

-s local1

-p 16000

-iqmsgsz 32

-iqmsgnum 2

-zr SQL

-zs 33554432

 

Had attached SYBASE ASE sp_sysmon output.

 

This test has been performed with 5 users.

 

I am not sure how to trace SYBASE IQ, but sp_iqconnection call gives 32 rows, which should mean 31 connections are made to SYBASE IQ db.

 

Please let us know how to further fine-tune SYBASE ASE and IQ database.

 

Thanks

Ramesh

↧

How to recover a suspect database ?

November 26, 2015, 2:35 am
$
0
0

Hello

my database is marked suspect.

I try sp_resetstatus mybase and reboot ASE, but the is still suspect.

 

version: Adaptive Server Enterprise/15.5/EBF 18164 SMP ESD#2/P/x86_64/Enterprise Linux/asear155/2514/64-bit/FBO/Wed Aug 25 11:17:26 2010

 

In the Errorlog:

Error: 2620, Severity: 21, State: 4

00:00:00000:00001:2015/11/26 10:37:50.05 server  The offset of the row number at offset 904 does not match the entry in the offset table of the following page: Page pointer = 0x0x151e3a000, pageno = 459332, status = 0x11, ptnid = 247109040, indexid = 0, level = 0.

00:00:00000:00001:2015/11/26 10:37:50.05 server  Database 'atelis' (dbid 4): Recovery failed. Check the ASE errorlog for further information as to the cause.

 

What can i do ?

↧

ASE on Solaris ZFS vs UFS

November 26, 2015, 12:11 pm
$
0
0

Hi Everybody,

 

 

Do you know if exist any benchmarks test between ASE on Solaris ZFS vs ASE on Solaris UFS?

 

 

Which one is the best option for an ASE server with HA configuration?

 

 

I found the SAP note 1878876, and there it says UFS is better choice for performance, but on HA Configuration is the same consideration?

 

 

Thanks in advance

 

Best Regards

 

Jairo

↧
↧

Prevent ASE from overriding trans logs ..

November 26, 2015, 7:17 am
$
0
0

Hi guys,

 

I'm using ASE since few month and have set up an environment with a PRD and a SDB.

With scripting I'm able to load the trans logs automatically into the SDB.

 

At the moment we are in the preparation phase of the EHP and got a strange issue, ASE has overridden a trans log, but it was created at a different time:

 

Dump history of PRD:

Nov 19 2015 11:25:18:826AM \\BACKUP\backup\PRD\log_archives\PRD.TRAN.20151119.112054.000

Nov 19 2015 11:25:22:826AM \\BACKUP\backup\PRD\log_archives\PRD.TRAN.20151119.112054.000

 

It looks like ASE had a problem at this moment and has written both after coming back into normal state ..

 

So SDB is missing the first created trans log and I had to rebuild the SDB from a fresh database backup.

But now I'm searching for a solution that this never happens again, but I don't find anything to prevent ASE from this issue!

 

Any idea?

 

 

Environment:

> select @@version> go
Adaptive Server Enterprise/16.0 GA PL02/EBF 22663 SMP/P/X64/Windows Server/ase160sp00pl02/3571/64-bit/FBO/Mon May 26 19:41:56 2014

 

1> sp_config_dump TRAN_DUMP_PRD
2> go
 optionname                               optionvalue
 -------------------------------------------- -------------------------------------
 stripe_dir                                  \\BACKUP\backup\PRD\log_archives
 compression                            101
 verify                                       header

 

Best regards,

Tobias

↧

Size of database dump(full)after the dump gets completed

October 8, 2015, 11:56 pm
$
0
0

Dear Experts,

 

In our environment we are dumping the database(full)to multiple stripes.After the entire dump gets completed, how can we find the size of the dump and where we can see the size exactly? Ideally, I am considering the size of dump that happened from backup server error file. Please let me know in case of any other ways to find out the dump size once the dump gets completed.

 

Regards,

Dilip Voora

↧

how to upgrade from ASE 12.5.4 to ASE 16.0?

November 25, 2015, 11:29 pm
$
0
0

is it  possible to be upgraded from 12.5.4 ESD#10 to 16.0 SP02 using dump/load? (IBM AIX)

without going through the intermediate step the ASE 15.x

↧
Search

Cross-platform conversion (12.5.3 --> 16.0)

December 2, 2015, 3:23 am
$
0
0

Hi,

 

I am new to Sybase and have been tasked with undertaking a proof of concept to move an existing Sybase installation (12.5.3) running on Solaris 9 to Sybase (16.0) running on Windows.

 

I have created the new Sybase instance and have managed to successfully convert a number of the user databases however I am getting an error (3422) on one particular database (see log attached).

 

Any suggestions? BTW - we do not have a current support contract so I am not able to raise an incident

 

Thanks,

Paul

↧
↧

Function-based index not work.

December 3, 2015, 6:44 pm
$
0
0

Hi,

 

function-based index not work on ASE15.7 sp132, its normal on sp100.

 

Here's the sample:

create table aaa ( a1 int , a2 int , a3 char(1000))

go

 

declare @cnt int

select @cnt = 1

while @cnt < 1001

     begin

          insert into aaa values ( @cnt , @cnt , "aaa")

          select @cnt = @cnt + 1

     end

go

 

update aaa set a3="aba" where a1 = 3

go

 

create index idx_aaa on aaa ( a1*a2 )

go

select * from aaa where a1*a2=333

go

==> will use index_aaa and logical reads is 2

 

 

create index idx_aaa1 on aaa ( substring(a3,2,1) )

go

select * from aaa where substring(a3,2,1)="b"

go

==> table scan and logical reads is 334

 

all these index will work on ASE15.7 sp100

 

any idea?

↧

No language handler installed

November 12, 2015, 7:28 pm
$
0
0

Hi,

 

Recently I installed ase16 and when I am trying to issue q command to an open server application, am getting this no language handler error.

 

But when I try the same with my actual prod setup it works perfectly.

 

I there anything I am going wrong here ?

 

Any reference on this would also be of great help.

↧

update Index Stats using Sampling

December 3, 2015, 11:30 am
$
0
0

Hi Team Good afternoon We currently have a JOB that runs UIS using sampling = 30 and is failing on a table that has 293412179 rows and one CI Error - the tempdb is full for the DB maintenance  a.c which is mapped to the DBA tempdb Can't allocate space for object 'temp worktable' in database 'DBA_tempdb' because 'system' segment is full/has no free extents. If you ran out of space in syslogs, dump the transaction log. Otherwise, use ALTER DATABASE to increase the size of the segment. When we run the same job using sampling = 10 it works fine. Question What are the risk associated with changing the sampling ? Will that impact the query plans for the app using that table ? -Sid

↧

"inserted" table not allowed in derived table query

December 7, 2015, 4:43 am
$
0
0

The "inserted" table not allowed in derived table query.

 

create trigger a on newtable

for insert

as

insert into othertable

select a,c from

(

    select distinct a,b,c from inserted

) dt

 

Msg 208, Level 16, State 1

Server 'DBS', Procedure 'a', Line 3

inserted not found. Specify owner.objectname or use sp_help to check whether the object exists (sp_help may produce lots of output).

 

Is this a documentated limitation in ASE ?

↧
↧

"configure size for data cache (tempdb_cache1) is not sufficient to handle the sort buffers (4005) requested. "

December 3, 2015, 9:57 am
$
0
0

Running a stored proc simultaneously a large number of times (> 150) I get the message

 

"The configure size for data cache (tempdb_cache1) is not sufficient to handle the sort buffers (4005) requested. Please reconfigure the data cache or sort buffers and try the command again.

Server:SDB_SQL Proc:GetEIDS Line:0 Severity:17 MessageNbr:1526"

 

I'm confused about the message what the correction action is.

 

I thought sort buffers were part of the procedure caches

 

    select * from monProcedureCacheModuleUsage

 

     InstanceID ModuleID    Active      HWM         NumPagesReused ModuleName

     ---------- ----------- ----------- ----------- -------------- ------------------------------

              0          11         463       68511              0 Sort

 

yet the message implies the the data cache is too small.

 

What's the relationship between data cache and sort buffers ?

 

We have

    "number of sort buffers" at 1000.

    "procedure cache size" at 19Gb

     tempdb_cache1           at 25Gb

 

sp_monitorconfig "all"

shows procedure cache hasn't run out but has reached 87% used (which is about the max we can deal with anyway)

 

Name                      Num_free    Num_active  Pct_act Max_Used    Reuse_cnt   Instance_Name

------------------------- ----------- ----------- ------- ----------- ----------- ------------------------------

procedure cache size          6182350     2521650  28.97      7487767       59073 NULL

 

 

From Simon Ogdens conversations, I'm inclined to think we need more procedure cache.

Or should I look at the data cache and sort buffers.

↧

sp_who ouput in ASE 16 sps01

December 7, 2015, 12:45 pm
$
0
0

In sp_who output it shows "compling" in cmd column . Is there any reason for it .

↧
Search

What am I waiting for again?

December 11, 2015, 12:38 pm
$
0
0
Timestamp  WaitEventID Descr                                          Waits   WaitTime
10:36:40        29 waiting for regular buffer read to complete            111     500
10:36:40        52 waiting for i/o on MASS initated by another task        12     200
10:36:40       150 waiting for a lock                                       3     100
10:36:40       171 waiting for CTLIB event to complete                    125     100

 

 

Timestamp  WaitEventID Descr                                          Waits   WaitTime
10:41:45       215 waiting on run queue after sleep                         1  315600
10:41:45        29 waiting for regular buffer read to complete            106     700
10:41:45        51 waiting for last i/o on MASS to complete                48     500
10:41:45        55 wait for i/o to finish after writing last log page      64     400
10:41:45       214 waiting on run queue after yield                        67     400
10:41:45        52 waiting for i/o on MASS initated by another task        26     200
10:41:45       124 wait for mass read to finish when getting page           7     100

 

 

Timestamp  WaitEventID Descr                                          Waits   WaitTime
10:47:53       215 waiting on run queue after sleep                         0  171500
10:47:53        29 waiting for regular buffer read to complete            874  158500
10:47:53       171 waiting for CTLIB event to complete                     25    2500

 

Above is the MDA output (deltas between observations) for a spid that I'm troubleshooting.  These are the wait events that have WaitTime > 0 for my spid.

 

My concern is waitevent 215 (waiting for run queue after sleep).  I understand that there is physical io going on here (events 29,51,52,55), but why the 315 seconds of waiting on single event of 215 (run queue), followed by 171 seconds (with waits=0, weird)?   This is a single SQL statement updating a TEXT field, and taking around 5 minutes (holding locks) to complete.  Are the physical io's of event 29 simply amplifying event 215 numbers?  Hoping to get an MDA mad scientist's input here

↧

ASE 16.0 Sp02 install over ASE 16.0

December 11, 2015, 6:31 am
$
0
0

    Good day,

I'm looking at ASE 16.0 SP02 install guide for Linux x64( readme  file in the tar file downloaded ) and there isn't a note  about loading instructions. So ,I assume that  it can be installed as the usual way EBFs are installed ?, i.e over an existing ASE 16.0 installation directory.

 

Thanks

Regards

Jose

↧
↧

Total beginner with Sybase

December 12, 2015, 8:58 pm
$
0
0

Hi!

I have just installed ASE 16 on Oracle linux 6.7. I managed to change the password for the user "sa". Is there a GUI tool that can be used to start server? How do I stop/start the database? I know about the startserver -f $SYBASE/ASE-16_0/install/RUN_SYBASE and RUN_SYBASE_BS but I would like a GUI tool for a start. I am trying to configure Commvault Simpana to work with Sybase.

Regards

↧

Sybase ASE transaction log's not clearing

December 15, 2015, 1:25 am
$
0
0

We are running Sybase ASE v.11

 

We are not able to clear out the transaction log's using a full DB backup. We have turned on 'trunc log on chkpt', and did one manual checkpoint using SQL advantage. Then we tried taking a complete DB backup, but the transaction log's are not clearing.

 

We also tried restarting the server and followed the above procedure, but yet it doesn't want to work. Is there anything we are missing out.

 

Kindly advise.


Regards,
Ghaffar Ahmed.

↧

How to access ASE remotely from a Linux server?

December 14, 2015, 12:33 pm
$
0
0

Hello Everybody,

 

I have a Linux server running RedHat 6.5 and would like to remotely access (and run queries) ASE that is running on a Windows server. I'd like to know which rpms (redhat package management) files I would need to install on a Linux server in order to accomplish this and where to get those rpms.

The version of ASE we have is ASE Enterprise Edition Windows x86-32bit version 12.5

Any help would be appreciated.

 

Thank you,

Arthur.

↧
Viewing all 685 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>