Hi! I've an old machine with zope 2.6 with an application based on ZClass. Now, updating 4000 record takes ages (more than an hour...), basically it is a propertysheet update + recatalog of every object of 4000. Can I have significative performance upgrade changing something in the code base? Some ZCatalog "slow" bug in 2.6 I could patch you know of? :) TIA
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 - --On 23. November 2006 17:07:39 +0100 Yuri <yurj@alfa.it> wrote:
Hi!
I've an old machine with zope 2.6 with an application based on ZClass. Now, updating 4000 record takes ages (more than an hour...), basically it is a propertysheet update + recatalog of every object of 4000.
Can I have significative performance upgrade changing something in the code base? Some ZCatalog "slow" bug in 2.6 I could patch you know of? :)
ZopeProfiler will tell you about bottlenecks. - -aj -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin) iD8DBQFFZce2CJIWIbr9KYwRAjMFAKCje2mf4XN4z/loynCUMo+MIW4cSwCfUnUA dH8pNs+4XK2yrzjJ+sfVWlQ= =1Uee -----END PGP SIGNATURE-----
Andreas Jung wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
- --On 23. November 2006 17:07:39 +0100 Yuri <yurj@alfa.it> wrote:
Hi!
I've an old machine with zope 2.6 with an application based on ZClass. Now, updating 4000 record takes ages (more than an hour...), basically it is a propertysheet update + recatalog of every object of 4000.
Can I have significative performance upgrade changing something in the code base? Some ZCatalog "slow" bug in 2.6 I could patch you know of? :)
ZopeProfiler will tell you about bottlenecks.
Doesn't work in zope 2.6 based on python 2.1 :( # more version.txt 1.7.2 for Zope 2.5.1 and 2.6.1, 2.7.3, 2.8, 2.9 Python 2.3.x, 2.4.x
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 - --On 24. November 2006 11:16:50 +0100 Yuri <yurj@alfa.it> wrote:
ZopeProfiler will tell you about bottlenecks.
Doesn't work in zope 2.6 based on python 2.1 :(
Try CallProfiler instead. - -aj - -- ZOPYX Ltd. & Co. KG - Charlottenstr. 37/1 - 72070 Tübingen - Germany Web: www.zopyx.com - Email: info@zopyx.com - Phone +49 - 7071 - 793376 E-Publishing, Python, Zope & Plone development, Consulting -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin) iD8DBQFFZsd3CJIWIbr9KYwRAkt2AKCssal5dASk5L2Hr5Hx25NcP73HlACfd8sR zwcPSLgP0kd/wyMVPCtV4oU= =lx7b -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 - --On 24. November 2006 11:20:39 +0100 Andreas Jung <lists@zopyx.com> wrote:
ZopeProfiler will tell you about bottlenecks.
Doesn't work in zope 2.6 based on python 2.1 :(
Try CallProfiler instead.
and *enable* subtransactions! - -aj -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin) iD8DBQFFZsfJCJIWIbr9KYwRAn4EAJ9NDvAQT4jOEcBSOEshEWz/I6EtEwCfepEK tPE8sMBxMQET/euEhCVAD5g= =FztJ -----END PGP SIGNATURE-----
Andreas Jung wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
- --On 24. November 2006 11:20:39 +0100 Andreas Jung <lists@zopyx.com> wrote:
ZopeProfiler will tell you about bottlenecks.
Doesn't work in zope 2.6 based on python 2.1 :(
Try CallProfiler instead.
and *enable* subtransactions!
When I have a transaction wich takes almost 1 sec, I don't think it will change anything :) I'll try CallProfiler :)
Yuri wrote:
When I have a transaction wich takes almost 1 sec, I don't think it will change anything :)
You might be surprised... Chris -- Simplistix - Content Management, Zope & Python Consulting - http://www.simplistix.co.uk
On 11/24/06, Yuri <yurj@alfa.it> wrote:
and *enable* subtransactions!
When I have a transaction wich takes almost 1 sec, I don't think it will change anything :) I'll try CallProfiler :)
Did you try? Subtransactions are not the same as transactions. They are an optimization, so try them! Don't just ditch the experts here. -- Martijn Pieters
Martijn Pieters wrote:
On 11/24/06, Yuri <yurj@alfa.it> wrote:
and *enable* subtransactions!
When I have a transaction wich takes almost 1 sec, I don't think it will change anything :) I'll try CallProfiler :)
Did you try? Subtransactions are not the same as transactions. They are an optimization, so try them! Don't just ditch the experts here.
ok, you got me :) with callprofiler I have: +0.0209 0.1213 1.4% +- inserimento +0.0461 0.0044 0.1% | +- standard_html_header +0.0479 0.0018 0.0% | | +- intestazione_html_standard +0.0518 0.0012 0.0% | | +- mappa_barra_loggato 0.0931 1.1% | | +0.1467 7.4849 85.1% | +- dafne_item_add 1.1709 13.3% | | | +1.3175 0.0027 0.0% | | +- proprietario +1.3245 0.0450 0.5% | | +- sonopubblico 0.1486 1.7% | | | +1.5181 0.0026 0.0% | | +- proprietario (2nd call) 1.3108 14.9% | | | +2.8316 0.0043 0.0% | | +- globale 4.6964 53.4% | | | +7.5323 0.0452 0.5% | | +- sonopubblico (2nd call) 0.1539 1.7% | | | +7.7324 0.0367 0.4% | +- invalida_cache +7.7704 0.0010 0.0% | +- standard_html_footer What does exactly take 4.6 sec? :-o
----- Original Message ----- From: "Yuri" <yurj@alfa.it> To: <zope@zope.org> Sent: Thursday, November 23, 2006 11:07 AM Subject: [Zope] help performance
Hi!
I've an old machine with zope 2.6 with an application based on ZClass. Now, updating 4000 record takes ages (more than an hour...), basically it is a propertysheet update + recatalog of every object of 4000.
Can I have significative performance upgrade changing something in the code base? Some ZCatalog "slow" bug in 2.6 I could patch you know of? :)
You could try different transaction subcommit threshold settings (Catalog - Advanced tab) to see if different settings have any beneficial impact. Jonathan
Jonathan wrote:
----- Original Message ----- From: "Yuri" <yurj@alfa.it> To: <zope@zope.org> Sent: Thursday, November 23, 2006 11:07 AM Subject: [Zope] help performance
Hi!
I've an old machine with zope 2.6 with an application based on ZClass. Now, updating 4000 record takes ages (more than an hour...), basically it is a propertysheet update + recatalog of every object of 4000.
Can I have significative performance upgrade changing something in the code base? Some ZCatalog "slow" bug in 2.6 I could patch you know of? :)
You could try different transaction subcommit threshold settings (Catalog - Advanced tab) to see if different settings have any beneficial impact.
Subtransactions are *Disabled :) *
participants (5)
-
Andreas Jung -
Chris Withers -
Jonathan -
Martijn Pieters -
Yuri