Skip to main content

Performance Problems after migration to AL32UTF8

Author: hschmidt@versorgungskammer.de (Hendrik Schmidt)

Hello, we have Uniface 9.7.0.4.x and Oracle DB 12.1.0.2. we migrated the database characterset from WE8ISO8859P15 to AL32UTF8 using Oracle's DMU ( Database Migration Assistant for Unicode). We set in the ASN’s $DEF_CHARSET = utf8 $SYS_CHARSET = utf8 $WIDE_CHAR_BEHAVIOR = Unicode we changed in the repository the character field Definition from Cxxx or C* to Wxxxx or W* or U*  and MUL to FUL eg: update fare.ucfield set u_intf = replace(u_intf, ‘C’, ‘W’) where u_vlab = ‘MR’ and u_dtyp = ‘S’ and u_intf like ‘C%’ and u_flab <> ‘U_VERSION’; update fare.ucfield set u_syn = replace(u_syn, ‘MUL’, ‘FUL’) where u_vlab = ‘MR’ and u_syn like ‘%MUL%’; commit; -> 1661 Zeilen aktualisiert. update fare.uxfield set INTERFACEMOD = replace(INTERFACEMOD, ‘C’, ‘W’) where ubase = ‘MR’ and INTERFACEMOD like ‘C%’; -> 5916 Zeilen aktualisiert. update fare.uxfield set SYNTAXMOD = replace(SYNTAXMOD, ‘MUL’, ‘FUL’) where ubase = ‘MR’ and SYNTAXMOD like ‘%MUL%’; -> 3486 Zeilen aktualisiert. We recompiled whole Application We can insert/save/query Unicode characters eg: Ҹҵ but the Performance is very bad. It seems that the Display / Rendering of Formsmasks is very slow ( ~ 5 tims slower as before) eg. we have normally a save and commit button in the masks -> if we press it, the save comes very fast, but to go back to the previous mask and display it, it needs very long. Any ideas? Someone know this Situation? many thanks in advanced Hendrik

Performance Problems after migration to AL32UTF8

Author: hschmidt@versorgungskammer.de (Hendrik Schmidt)

Hello, we have Uniface 9.7.0.4.x and Oracle DB 12.1.0.2. we migrated the database characterset from WE8ISO8859P15 to AL32UTF8 using Oracle's DMU ( Database Migration Assistant for Unicode). We set in the ASN’s $DEF_CHARSET = utf8 $SYS_CHARSET = utf8 $WIDE_CHAR_BEHAVIOR = Unicode we changed in the repository the character field Definition from Cxxx or C* to Wxxxx or W* or U*  and MUL to FUL eg: update fare.ucfield set u_intf = replace(u_intf, ‘C’, ‘W’) where u_vlab = ‘MR’ and u_dtyp = ‘S’ and u_intf like ‘C%’ and u_flab <> ‘U_VERSION’; update fare.ucfield set u_syn = replace(u_syn, ‘MUL’, ‘FUL’) where u_vlab = ‘MR’ and u_syn like ‘%MUL%’; commit; -> 1661 Zeilen aktualisiert. update fare.uxfield set INTERFACEMOD = replace(INTERFACEMOD, ‘C’, ‘W’) where ubase = ‘MR’ and INTERFACEMOD like ‘C%’; -> 5916 Zeilen aktualisiert. update fare.uxfield set SYNTAXMOD = replace(SYNTAXMOD, ‘MUL’, ‘FUL’) where ubase = ‘MR’ and SYNTAXMOD like ‘%MUL%’; -> 3486 Zeilen aktualisiert. We recompiled whole Application We can insert/save/query Unicode characters eg: Ҹҵ but the Performance is very bad. It seems that the Display / Rendering of Formsmasks is very slow ( ~ 5 tims slower as before) eg. we have normally a save and commit button in the masks -> if we press it, the save comes very fast, but to go back to the previous mask and display it, it needs very long. Any ideas? Someone know this Situation? many thanks in advanced Hendrik

Hello Hendrik, I was just about to write a similar problem and I saw your post. We migrated our application from Uniface 9.6.08 to 9.7.04 and detected that a screen that was being opened immediately now takes about 6-7 seconds to be opened. Debugging we have seen that it occurs also when the screen is drawn through the command Display in the form. Nothing has changed in the database configuration or model. Any ideas? Did you solve the problem? Thanks in advance and regards, José Manuel


Author: jverdugocall (jverdugocall@csc.com)

Performance Problems after migration to AL32UTF8

Author: hschmidt@versorgungskammer.de (Hendrik Schmidt)

Hello, we have Uniface 9.7.0.4.x and Oracle DB 12.1.0.2. we migrated the database characterset from WE8ISO8859P15 to AL32UTF8 using Oracle's DMU ( Database Migration Assistant for Unicode). We set in the ASN’s $DEF_CHARSET = utf8 $SYS_CHARSET = utf8 $WIDE_CHAR_BEHAVIOR = Unicode we changed in the repository the character field Definition from Cxxx or C* to Wxxxx or W* or U*  and MUL to FUL eg: update fare.ucfield set u_intf = replace(u_intf, ‘C’, ‘W’) where u_vlab = ‘MR’ and u_dtyp = ‘S’ and u_intf like ‘C%’ and u_flab <> ‘U_VERSION’; update fare.ucfield set u_syn = replace(u_syn, ‘MUL’, ‘FUL’) where u_vlab = ‘MR’ and u_syn like ‘%MUL%’; commit; -> 1661 Zeilen aktualisiert. update fare.uxfield set INTERFACEMOD = replace(INTERFACEMOD, ‘C’, ‘W’) where ubase = ‘MR’ and INTERFACEMOD like ‘C%’; -> 5916 Zeilen aktualisiert. update fare.uxfield set SYNTAXMOD = replace(SYNTAXMOD, ‘MUL’, ‘FUL’) where ubase = ‘MR’ and SYNTAXMOD like ‘%MUL%’; -> 3486 Zeilen aktualisiert. We recompiled whole Application We can insert/save/query Unicode characters eg: Ҹҵ but the Performance is very bad. It seems that the Display / Rendering of Formsmasks is very slow ( ~ 5 tims slower as before) eg. we have normally a save and commit button in the masks -> if we press it, the save comes very fast, but to go back to the previous mask and display it, it needs very long. Any ideas? Someone know this Situation? many thanks in advanced Hendrik

Hello all, many thanks for the updates. After some other work, we want to test the ideas. I think the last one with connector Settings seems to be a very hot Point.   Regards Hendrik


Author: Hendrik Schmidt (hschmidt@versorgungskammer.de)

Performance Problems after migration to AL32UTF8

Author: hschmidt@versorgungskammer.de (Hendrik Schmidt)

Hello, we have Uniface 9.7.0.4.x and Oracle DB 12.1.0.2. we migrated the database characterset from WE8ISO8859P15 to AL32UTF8 using Oracle's DMU ( Database Migration Assistant for Unicode). We set in the ASN’s $DEF_CHARSET = utf8 $SYS_CHARSET = utf8 $WIDE_CHAR_BEHAVIOR = Unicode we changed in the repository the character field Definition from Cxxx or C* to Wxxxx or W* or U*  and MUL to FUL eg: update fare.ucfield set u_intf = replace(u_intf, ‘C’, ‘W’) where u_vlab = ‘MR’ and u_dtyp = ‘S’ and u_intf like ‘C%’ and u_flab <> ‘U_VERSION’; update fare.ucfield set u_syn = replace(u_syn, ‘MUL’, ‘FUL’) where u_vlab = ‘MR’ and u_syn like ‘%MUL%’; commit; -> 1661 Zeilen aktualisiert. update fare.uxfield set INTERFACEMOD = replace(INTERFACEMOD, ‘C’, ‘W’) where ubase = ‘MR’ and INTERFACEMOD like ‘C%’; -> 5916 Zeilen aktualisiert. update fare.uxfield set SYNTAXMOD = replace(SYNTAXMOD, ‘MUL’, ‘FUL’) where ubase = ‘MR’ and SYNTAXMOD like ‘%MUL%’; -> 3486 Zeilen aktualisiert. We recompiled whole Application We can insert/save/query Unicode characters eg: Ҹҵ but the Performance is very bad. It seems that the Display / Rendering of Formsmasks is very slow ( ~ 5 tims slower as before) eg. we have normally a save and commit button in the masks -> if we press it, the save comes very fast, but to go back to the previous mask and display it, it needs very long. Any ideas? Someone know this Situation? many thanks in advanced Hendrik

Hello Hendrik Have you tried the connector option uselike from the Uniface driver for Oracle: Use Use uselike only if you encounter performance problems when working with Unicode data and the database is set up with Unicode character sets (UTF8, UTF16, and so on), because LIKE calculates strings using characters as defined by the input character set. Description By default, the connector invokes the Oracle LIKEC operator when the where clause is used in SQL queries with Unicode data ($DEF_CHARSET =UTF8 or when the field has a W interface). In some configurations, this may lead to performance problems.   We introduced because likec from Oracle has performance  issues Peter


Author: PBeugel (peter.beugel@uniface.com)

Performance Problems after migration to AL32UTF8

Author: hschmidt@versorgungskammer.de (Hendrik Schmidt)

Hello, we have Uniface 9.7.0.4.x and Oracle DB 12.1.0.2. we migrated the database characterset from WE8ISO8859P15 to AL32UTF8 using Oracle's DMU ( Database Migration Assistant for Unicode). We set in the ASN’s $DEF_CHARSET = utf8 $SYS_CHARSET = utf8 $WIDE_CHAR_BEHAVIOR = Unicode we changed in the repository the character field Definition from Cxxx or C* to Wxxxx or W* or U*  and MUL to FUL eg: update fare.ucfield set u_intf = replace(u_intf, ‘C’, ‘W’) where u_vlab = ‘MR’ and u_dtyp = ‘S’ and u_intf like ‘C%’ and u_flab <> ‘U_VERSION’; update fare.ucfield set u_syn = replace(u_syn, ‘MUL’, ‘FUL’) where u_vlab = ‘MR’ and u_syn like ‘%MUL%’; commit; -> 1661 Zeilen aktualisiert. update fare.uxfield set INTERFACEMOD = replace(INTERFACEMOD, ‘C’, ‘W’) where ubase = ‘MR’ and INTERFACEMOD like ‘C%’; -> 5916 Zeilen aktualisiert. update fare.uxfield set SYNTAXMOD = replace(SYNTAXMOD, ‘MUL’, ‘FUL’) where ubase = ‘MR’ and SYNTAXMOD like ‘%MUL%’; -> 3486 Zeilen aktualisiert. We recompiled whole Application We can insert/save/query Unicode characters eg: Ҹҵ but the Performance is very bad. It seems that the Display / Rendering of Formsmasks is very slow ( ~ 5 tims slower as before) eg. we have normally a save and commit button in the masks -> if we press it, the save comes very fast, but to go back to the previous mask and display it, it needs very long. Any ideas? Someone know this Situation? many thanks in advanced Hendrik

Hi, I generally try to use this tool when I get into some perfomance issues since all process info related to the Uniface process will be trapped; https://docs.microsoft.com/en-us/sysinternals/downloads/procmon Also, you didn't say anyting about the Oracle driver you're using as well as the Oracle client... I've had instances in the past where there was a version mis-match between either the Uniface driver (rare) or the DBMS client sw. Knut


Author: Knut (knut.dybendahl@gmail.com)
Terms & ConditionsAccessibility statement