[Bug] - CB utility analyzedb does not work

[vsbace]

Apache Cloudberry version

We are trying to test Apache Cloudberry 2.1.0 on Astra Linux 1.8.2 (Debian 12 based)

What happened

Hello,

CB utility analyzedb does not work
gpadmin@qc1:/data/cbdb/master-1/gpseg-1$ analyzedb -d dwh -p 10 20260826:18:17:57:163972 analyzedb:qc1:gpadmin-[INFO]:-Starting analyzedb with args: -d dwh -p 10 20260826:18:17:57:163972 analyzedb:qc1:gpadmin-[INFO]:-Getting and verifying input tables and materialized views... 20260826:18:17:58:163972 analyzedb:qc1:gpadmin-[INFO]:-Checking for tables with stale stats... 20260826:18:17:58:163972 analyzedb:qc1:gpadmin-[ERROR]:-'ascii' codec can't encode characters in position 13-19: ordinal not in range(128) Traceback (most recent call last): File "/usr/local/cloudberry-db/bin/analyzedb", line 377, in execute heap_partitions = get_heap_tables_set(self.conn, input_tables_set) # set((schema1,table1), ...]) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/cloudberry-db/bin/analyzedb", line 1005, in get_heap_tables_set oid_str = get_oid_str(input_tables_set) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/cloudberry-db/bin/analyzedb", line 977, in get_oid_str return ','.join(map((lambda x: regclass_schema_tbl(x[0], x[1])), table_list)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/cloudberry-db/bin/analyzedb", line 977, in <lambda> return ','.join(map((lambda x: regclass_schema_tbl(x[0], x[1])), table_list)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/cloudberry-db/bin/analyzedb", line 985, in regclass_schema_tbl return "to_regclass('%s')" % (pg.escape_string(schema_tbl)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ UnicodeEncodeError: 'ascii' codec can't encode characters in position 13-19: ordinal not in range(128) 20260826:18:17:58:163972 analyzedb:kvd-dwh-qc1-c1:gpadmin-[CRITICAL]:-analyzedb failed. (Reason=''ascii' codec can't encode characters in position 13-19: ordinal not in range(128)') exiting...

What you think should happen instead

No response

How to reproduce

I am not sure about this problem.

Operating System

Astra Linux 1.8.2 (Debian 12 based)

Anything else

No response

Are you willing to submit PR?

• Yes, I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct.

[tuhaihe]

Thanks for the report. This is a real bug in analyzedb and it is not specific to Astra Linux — it hits any database that has a schema, table or column whose name is not pure ASCII, and it is still present on main.

I have opened a draft PR with a fix: #1938

Could you apply it and run analyzedb -d dwh -p 10 again?

Either way, please also post the following, so we can check whether anything else in the same area needs fixing — analyzedb writes its state and report files using the locale's encoding, and those files contain table names:

• output of locale
• SHOW client_encoding; and SHOW server_encoding; in the dwh database

[leborchuk]

By the way, how do you get binaries for work on Astra Linux 1.8.2 ? Maybe we could easily add support Astra Linux since we already has deb-based binaries

[vsbace]

Hi,

If you will add a Astra Linux support it will very good for us.

All times yarly i have had compiled Cloudberry by self for existing OS (Debian 12, Ubuntu 22, Astra Linux 1.8).

First step - i was installed all required packages on all cluster hosts:
apt install bison \ bzip2 \ cmake \ curl \ flex \ gcc \ g++ \ iproute2 \ iputils-ping \ locales \ libapr1-dev \ libbz2-dev \ libcurl4-gnutls-dev \ libevent-dev \ libkrb5-dev \ libipc-run-perl \ libldap2-dev \ libpam0g-dev \ libprotobuf-dev \ libreadline-dev \ libssl-dev \ libuv1-dev \ liblz4-dev \ libxerces-c-dev \ libxml2-dev \ libyaml-dev \ libzstd-dev \ libperl-dev \ make \ pkg-config \ protobuf-compiler \ python3-dev \ python3-pip \ python3-setuptools \ rsync

Second step -i was compiled the Cloudberry sources by the next commands
./configure --prefix=/usr/local/cloudberry-db \ --enable-debug \ --enable-debug-extensions \ --enable-ic-proxy \ --enable-orca \ --enable-pxf \ --with-gssapi \ --with-ldap \ --with-libxml \ --with-lz4 \ --with-pam \ --with-perl \ --with-python \ --with-ssl=openssl \ --with-uuid=e2fs make -j8 install

And on final step i was prepared DEB package for integrating with CI CD systems.

[vsbace]

Thanks for the report. This is a real bug in analyzedb and it is not specific to Astra Linux — it hits any database that has a schema, table or column whose name is not pure ASCII, and it is still present on main.

I have opened a draft PR with a fix: #1938

Could you apply it and run analyzedb -d dwh -p 10 again?

Either way, please also post the following, so we can check whether anything else in the same area needs fixing — analyzedb writes its state and report files using the locale's encoding, and those files contain table names:

• output of locale
• SHOW client_encoding; and SHOW server_encoding; in the dwh database

Hello,

`[root@qc1:~]# locale
LANG=en_US.UTF-8
LANGUAGE=
LC_CTYPE="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_PAPER="en_US.UTF-8"
LC_NAME="en_US.UTF-8"
LC_ADDRESS="en_US.UTF-8"
LC_TELEPHONE="en_US.UTF-8"
LC_MEASUREMENT="en_US.UTF-8"
LC_IDENTIFICATION="en_US.UTF-8"
LC_ALL=

gpadmin@qc1:~$ psql -d dwh
psql (14.4, server 14.4)
Type "help" for help.

dwh=# SHOW server_encoding;
server_encoding

UTF8
(1 row)

dwh=# SHOW client_encoding;
client_encoding

UTF8
(1 row)

`

[tuhaihe]

Thanks — UTF-8 on both the database and the OS side is good news.

One small thing: that locale output is from the root shell, but analyzedb runs as gpadmin. Could you post sudo -iu gpadmin locale as well? analyzedb writes its state and report files using the locale encoding of the account it runs under, and those files contain the table names, so that is the one that matters here.

And whenever you get a chance — with #1938 applied, what does analyzedb -d dwh -p 10 do now?

[vsbace]

Thanks — UTF-8 on both the database and the OS side is good news.

One small thing: that locale output is from the root shell, but analyzedb runs as gpadmin. Could you post sudo -iu gpadmin locale as well? analyzedb writes its state and report files using the locale encoding of the account it runs under, and those files contain the table names, so that is the one that matters here.

And whenever you get a chance — with #1938 applied, what does analyzedb -d dwh -p 10 do now?

`gpadmin@qc1:~$ id
uid=999(gpadmin) gid=999(gpadmin) groups=999(gpadmin),27(sudo),42(shadow)

gpadmin@qc1:~$ locale
LANG=en_US.UTF-8
LANGUAGE=
LC_CTYPE="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_PAPER="en_US.UTF-8"
LC_NAME="en_US.UTF-8"
LC_ADDRESS="en_US.UTF-8"
LC_TELEPHONE="en_US.UTF-8"
LC_MEASUREMENT="en_US.UTF-8"
LC_IDENTIFICATION="en_US.UTF-8"
LC_ALL=
`
I will plan to apply patch and test this utility.
Thank you for the fast response.