Sybase NNTP forums - End Of Life (EOL)

The NNTP forums from Sybase - forums.sybase.com - are now closed.

All new questions should be directed to the appropriate forum at the SAP Community Network (SCN).

Individual products have links to the respective forums on SCN, or you can go to SCN and search for your product in the search box (upper right corner) to find your specific developer center.

High System CPU% utilization on Sybase ASE Servers -- compared to user cpu%

5 posts in Performance and Tuning Last posting was on 2011-11-27 05:05:52.0Z
NJS Posted on 2011-11-18 17:55:30.0Z
From: NJS <isanth@optonline.net>
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:7.0.1) Gecko/20110929 Thunderbird/7.0.1
MIME-Version: 1.0
Newsgroups: sybase.public.ase.performance+tuning
Subject: High System CPU% utilization on Sybase ASE Servers -- compared to user cpu%
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4ec69c12@forums-1-dub>
Date: 18 Nov 2011 09:55:30 -0800
X-Trace: forums-1-dub 1321638930 10.22.241.152 (18 Nov 2011 09:55:30 -0800)
X-Original-Trace: 18 Nov 2011 09:55:30 -0800, vip152.sybase.com
Lines: 10
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.performance+tuning:13018
Article PK: 91472

We are noticing some pretty light ASE servers (12.5.4 ESD#10) utilizing
more system user cpU% than user cpu% (when looking at sar/top/vmstat).

Is there any issues with Sybase's 512 bytes packet size which causes
excessive context_switching, which could spike system cpu% ?

I will keep digging into the low-level linux resources and correlate
that with sp_sysmon output, but any pointers are very much appreciated.

thanks in advance


NJS Posted on 2011-11-23 13:53:13.0Z
From: NJS <isanth@optonline.net>
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:7.0.1) Gecko/20110929 Thunderbird/7.0.1
MIME-Version: 1.0
Newsgroups: sybase.public.ase.performance+tuning
Subject: Re: High System CPU% utilization on Sybase ASE Servers -- compared touser cpu%
References: <4ec69c12@forums-1-dub>
In-Reply-To: <4ec69c12@forums-1-dub>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4eccfac9$1@forums-1-dub>
Date: 23 Nov 2011 05:53:13 -0800
X-Trace: forums-1-dub 1322056393 10.22.241.152 (23 Nov 2011 05:53:13 -0800)
X-Original-Trace: 23 Nov 2011 05:53:13 -0800, vip152.sybase.com
Lines: 24
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.performance+tuning:13019
Article PK: 91469

Never mind. this seems a bug which is fixed somewhere close to 12.5.4
ESD#10 and ASE 15.x does not exhibit this problem.

RPSC (runnable process search count) is an engima on
blocking/non-blocking of CPU, and this can eat into the CPU needed by
other processes including other engines on the linux host.

RHEL6 seems to have some smart tools to do some /proc based
metrics/analysis (perf tool, system_tap etc..), so hopefully, we can get
some better accounting on cpu consumption (user vs system).

On 11/18/2011 12:55 PM, NJS wrote:
> We are noticing some pretty light ASE servers (12.5.4 ESD#10) utilizing
> more system user cpU% than user cpu% (when looking at sar/top/vmstat).
>
> Is there any issues with Sybase's 512 bytes packet size which causes
> excessive context_switching, which could spike system cpu% ?
>
> I will keep digging into the low-level linux resources and correlate
> that with sp_sysmon output, but any pointers are very much appreciated.
>
> thanks in advance


Manish Negandhi [TeamSybase] Posted on 2011-11-24 17:14:47.0Z
From: "Manish Negandhi [TeamSybase]" <nospam_negandhi.manish@gmail.com>
Newsgroups: sybase.public.ase.performance+tuning
References: <4ec69c12@forums-1-dub> <4eccfac9$1@forums-1-dub>
Subject: Re: High System CPU% utilization on Sybase ASE Servers -- compared touser cpu%
Lines: 35
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2900.3664
X-RFC2646: Format=Flowed; Response
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3664
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4ece7b87@forums-1-dub>
Date: 24 Nov 2011 09:14:47 -0800
X-Trace: forums-1-dub 1322154887 10.22.241.152 (24 Nov 2011 09:14:47 -0800)
X-Original-Trace: 24 Nov 2011 09:14:47 -0800, vip152.sybase.com
X-Authenticated-User: teamsybase
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.performance+tuning:13020
Article PK: 91470

I have seen similar problem in my Linux environment, would you mind giving
case or CR number for this ?

-Manish Negandhi
[TeamSybase]

"NJS" <isanth@optonline.net> wrote in message
news:4eccfac9$1@forums-1-dub...
> Never mind. this seems a bug which is fixed somewhere close to 12.5.4
> ESD#10 and ASE 15.x does not exhibit this problem.
>
> RPSC (runnable process search count) is an engima on blocking/non-blocking
> of CPU, and this can eat into the CPU needed by other processes including
> other engines on the linux host.
>
> RHEL6 seems to have some smart tools to do some /proc based
> metrics/analysis (perf tool, system_tap etc..), so hopefully, we can get
> some better accounting on cpu consumption (user vs system).
>
>
> On 11/18/2011 12:55 PM, NJS wrote:
>> We are noticing some pretty light ASE servers (12.5.4 ESD#10) utilizing
>> more system user cpU% than user cpu% (when looking at sar/top/vmstat).
>>
>> Is there any issues with Sybase's 512 bytes packet size which causes
>> excessive context_switching, which could spike system cpu% ?
>>
>> I will keep digging into the low-level linux resources and correlate
>> that with sp_sysmon output, but any pointers are very much appreciated.
>>
>> thanks in advance
>


NJS Posted on 2011-11-26 00:49:09.0Z
From: NJS <isanth@optonline.net>
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:7.0.1) Gecko/20110929 Thunderbird/7.0.1
MIME-Version: 1.0
Newsgroups: sybase.public.ase.performance+tuning
Subject: Re: High System CPU% utilization on Sybase ASE Servers -- compared touser cpu%
References: <4ec69c12@forums-1-dub> <4eccfac9$1@forums-1-dub> <4ece7b87@forums-1-dub>
In-Reply-To: <4ece7b87@forums-1-dub>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4ed03785$1@forums-1-dub>
Date: 25 Nov 2011 16:49:09 -0800
X-Trace: forums-1-dub 1322268549 10.22.241.152 (25 Nov 2011 16:49:09 -0800)
X-Original-Trace: 25 Nov 2011 16:49:09 -0800, vip152.sybase.com
Lines: 45
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.performance+tuning:13021
Article PK: 91471

I don't have the CR number. The Sybase support person I spoke to
indicated a CR number which was more for solaris than Linux (and seemed
to match a case with the brk() call ), but it kind of matched with the
problem I had. ASE 15, ASE 15.5 and 12.5.4 ESD#10 and above are immune
to the problem, so it might be something to do with excessive hold-up of
the cpu during the select polling call.

On 11/24/2011 12:14 PM, Manish Negandhi [TeamSybase] wrote:
> I have seen similar problem in my Linux environment, would you mind giving
> case or CR number for this ?
>
> -Manish Negandhi
> [TeamSybase]
>
>
> "NJS"<isanth@optonline.net> wrote in message
> news:4eccfac9$1@forums-1-dub...
>> Never mind. this seems a bug which is fixed somewhere close to 12.5.4
>> ESD#10 and ASE 15.x does not exhibit this problem.
>>
>> RPSC (runnable process search count) is an engima on blocking/non-blocking
>> of CPU, and this can eat into the CPU needed by other processes including
>> other engines on the linux host.
>>
>> RHEL6 seems to have some smart tools to do some /proc based
>> metrics/analysis (perf tool, system_tap etc..), so hopefully, we can get
>> some better accounting on cpu consumption (user vs system).
>>
>>
>> On 11/18/2011 12:55 PM, NJS wrote:
>>> We are noticing some pretty light ASE servers (12.5.4 ESD#10) utilizing
>>> more system user cpU% than user cpu% (when looking at sar/top/vmstat).
>>>
>>> Is there any issues with Sybase's 512 bytes packet size which causes
>>> excessive context_switching, which could spike system cpu% ?
>>>
>>> I will keep digging into the low-level linux resources and correlate
>>> that with sp_sysmon output, but any pointers are very much appreciated.
>>>
>>> thanks in advance
>>
>
>


Manish Negandhi [TeamSybase] Posted on 2011-11-27 05:05:52.0Z
From: "Manish Negandhi [TeamSybase]" <nospam_negandhi.manish@gmail.com>
Newsgroups: sybase.public.ase.performance+tuning
References: <4ec69c12@forums-1-dub> <4eccfac9$1@forums-1-dub> <4ece7b87@forums-1-dub> <4ed03785$1@forums-1-dub>
Subject: Re: High System CPU% utilization on Sybase ASE Servers -- compared touser cpu%
Lines: 58
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2900.3664
X-RFC2646: Format=Flowed; Response
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.3664
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4ed1c530$1@forums-1-dub>
Date: 26 Nov 2011 21:05:52 -0800
X-Trace: forums-1-dub 1322370352 10.22.241.152 (26 Nov 2011 21:05:52 -0800)
X-Original-Trace: 26 Nov 2011 21:05:52 -0800, vip152.sybase.com
X-Authenticated-User: teamsybase
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.performance+tuning:13022
Article PK: 91474

Thanks for your help, I got it

Manish Negandhi
[TeamSybase]

"NJS" <isanth@optonline.net> wrote in message
news:4ed03785$1@forums-1-dub...
>I don't have the CR number. The Sybase support person I spoke to indicated
>a CR number which was more for solaris than Linux (and seemed to match a
>case with the brk() call ), but it kind of matched with the problem I had.
>ASE 15, ASE 15.5 and 12.5.4 ESD#10 and above are immune to the problem, so
>it might be something to do with excessive hold-up of the cpu during the
>select polling call.
>
>
> On 11/24/2011 12:14 PM, Manish Negandhi [TeamSybase] wrote:
>> I have seen similar problem in my Linux environment, would you mind
>> giving
>> case or CR number for this ?
>>
>> -Manish Negandhi
>> [TeamSybase]
>>
>>
>> "NJS"<isanth@optonline.net> wrote in message
>> news:4eccfac9$1@forums-1-dub...
>>> Never mind. this seems a bug which is fixed somewhere close to 12.5.4
>>> ESD#10 and ASE 15.x does not exhibit this problem.
>>>
>>> RPSC (runnable process search count) is an engima on
>>> blocking/non-blocking
>>> of CPU, and this can eat into the CPU needed by other processes
>>> including
>>> other engines on the linux host.
>>>
>>> RHEL6 seems to have some smart tools to do some /proc based
>>> metrics/analysis (perf tool, system_tap etc..), so hopefully, we can get
>>> some better accounting on cpu consumption (user vs system).
>>>
>>>
>>> On 11/18/2011 12:55 PM, NJS wrote:
>>>> We are noticing some pretty light ASE servers (12.5.4 ESD#10) utilizing
>>>> more system user cpU% than user cpu% (when looking at sar/top/vmstat).
>>>>
>>>> Is there any issues with Sybase's 512 bytes packet size which causes
>>>> excessive context_switching, which could spike system cpu% ?
>>>>
>>>> I will keep digging into the low-level linux resources and correlate
>>>> that with sp_sysmon output, but any pointers are very much appreciated.
>>>>
>>>> thanks in advance
>>>
>>
>>
>