Pxc 4469 sst clone #1961

Tusamarco · 2024-10-07T17:08:10Z

Adding preview POC for the Clone method for SST.
The script is based on the one from codership but then need to refactor almost in full to have it working with PXC.
birdeye flow

#              ┌──────────────────┐                      
#              │  Joiner starts   │                      
#              └────────┬─────────┘                      
#              ┌────────▼─────────┐                      
#              │open NetCat listnr│                      
#              └────────┬─────────┘                      
#              ┌────────▼─────────┐                      
#              │Send message to Dn│                      
#              └────────┬─────────┘                      
#              ┌────────▼─────────┐                      
#              │Donor decide IST  │                      
#              │or SST. Send msg  │                      
#              │through NC and    │                      
#              │waits for Joiner  │                      
#              │Clone instance    │                      
#              └────────┬─────────┘                      
#              ┌────────▼─────────┐                      
#              │Joiner get IST    │                      
#              │or SST.If IST     │                      
#              │Bypass all and    │                      
#              │Wait for IST.     │                      
#              │SST Start clone   │                      
#              │Instance and wait │                      
#              │for donor to clone│                      
#              └─────────┬────────┘                      
#              ┌─────────▼─────────────┐                 
#              │Clone process is       │                 
#              │reported.              │                 
#              │When done Dn waits     │                 
#              │Joiner close instance  │                 
#              │And performs 3 restarts│                 
#              └──────────┬────────────┘                 
#              ┌──────────▼──────────────┐               
#              │1) To fix dictionary     │               
#              │2) To recover position   │               
#              │3) Final cleanup         │               
#              └──────────┬──────────────┘               
#              ┌──────────▼──────────────┐               
#              │Send final ready signal  │               
#              │Waits for IST            │               
#              └─────────────────────────┘

This version has a patch to go straight to ist so is not fully functional unless removing that bypass

SSL is NOT supported

bug. Wrong comma

fixing config missed wsrep allowed methods

…tcat is not running

it-percona-cla · 2024-10-07T17:08:15Z

All committers have signed the CLA.

Tusamarco · 2024-10-07T17:18:36Z

I have a question...
the two file wsrep_clone and wsrep_common_clone are not copied over when I do make install where is that defined?
@kamil-holubicki @venkatesh-prasad-v

Need to ad extension or make script will not identify them correctly

…the user Some edit/cleanup as suggested by Kamil

kamil-holubicki · 2024-11-12T17:00:37Z

mysql-test/suite/galera/include/galera_st_kill_slave_ddl.inc

@@ -1,7 +1,7 @@
 --echo Performing State Transfer on a server that has been killed and restarted
 --echo while a DDL was in progress on it

--source include/have_debug.inc
+#source include/have_debug.inc


I think it will cause another tests to fail

mysql-test/suite/galera/t/galera_sst_clone.cnf

kamil-holubicki · 2024-11-12T17:05:01Z

mysql-test/suite/galera/t/galera_sst_clone.cnf

+[sst]
+netcat_port=4442
+wsrep-debug=true
+clone_instance_port=4444


I'm not sure we need clone_instance_port parameter at all. Please see following comments in sst script for justification/reasoning.

we do.
This may go through firewall in real life, and the use of port and ranges can be regulated. So better to keep the flexibility to define ports from the beginning.

We already have a way to configure this port.
See my other comment about wsrep_sst_receive_address

I think we should remove these new params from cnf after rework

mysql-test/suite/galera/t/galera_sst_clone.test

scripts/wsrep_sst_common_clone.sh

scripts/wsrep_sst_clone.sh

kamil-holubicki · 2024-11-12T19:01:22Z

scripts/mysql_system_users.sql

+
+
+
+GRANT ALTER, CREATE, SELECT, INSERT ON PERCONA_SCHEMA.xtrabackup_history TO 'mysql.pxc.sst.role'@localhost;


Why do we need this?

No Idea it was there, not related to changes I have add.

kamil-holubicki · 2024-11-12T19:02:00Z

sql/wsrep_sst.cc

@@ -1230,6 +1230,8 @@ int wsrep_remove_sst_user(bool initialize_thread) {
                        nullptr,
                        "SET SESSION lock_wait_timeout = 1;",
                        nullptr,
+                        "REVOKE mysql.pxc.sst.role FROM 'mysql.pxc.sst.user'@'localhost';",


Why is this needed?

We discuss this.
The correct way to remove the user when having a role, is to revoke the role and then drop the user.
Just dropping the user will work, but the role cleanup is delegated to MySQL, and as we have identified with mtr tests on very fast machines this looks like to be asynchronous.
So to avoid problems and to be on the safe side, we should operate following the right execution. Role first then drop user.

scripts/mysql_system_users.sql

sql/wsrep_sst.cc

scripts/wsrep_sst_common_clone.sh

kamil-holubicki · 2024-11-12T19:08:36Z

As we spoke today, we need to test it with encrypted tables and keyring component/plugin enabled

- removed the need to perform the restart to clean up the final instance. - use wsrep-sst-receive-address to get IP and port definition - use one port for Netcat and MySQL - some code cleanup - Add message suppression when reporting clone Status %. It will be reported only if different from previous value.

Tusamarco · 2024-11-20T14:58:26Z

@kamil-holubicki

==============================================================================
                  TEST NAME                       RESULT  TIME (ms) COMMENT
------------------------------------------------------------------------------
[ 33%] galera.galera_sst_clone 'enc_off'         [ pass ]  28381
[ 66%] galera.galera_sst_clone 'enc_on'          [ pass ]  28789
[100%] shutdown_report                           [ pass ]       
------------------------------------------------------------------------------
The servers were restarted 1 times
The servers were reinitialized 0 times
Spent 57.170 of 168 seconds executing testcases

Completed: All 3 tests were successful.

If you outline what kind of operation are you expecting to test when talking about data encryption I can see how clone deal with it. Also if think this can be step 2

venkatesh-prasad-v · 2024-11-22T05:49:42Z

scripts/wsrep_sst_clone.sh

@@ -0,0 +1,1113 @@
+#!/usr/bin/env bash


All the wsrep_sst*.sh files have 775 as the default file permission. Can you also do the same for wsrep_sst_clone.sh as well?

-rw-rw-r-- 1 venki venki 42263 Nov 21 22:17 wsrep_sst_clone.sh -rwxrwxr-x 1 venki venki 35144 Jan 26 2024 wsrep_sst_common.sh* -rwxrwxr-x 1 venki venki 23466 Jun 28 2023 wsrep_sst_rsync.sh* -rwxrwxr-x 1 venki venki 102235 Mar 7 2024 wsrep_sst_xtrabackup-v2.sh*

I can also if I was expecting to have these thing standardized by the building merging process. but ok will do

venkatesh-prasad-v · 2024-11-22T05:56:10Z

mysql-test/suite/galera/r/galera_sst_clone.result

+CALL mtr.add_suppression("Found an entry in the 'role_edges' table with unknown authorization ID '`mysql.pxc.sst.user`@`localhost`'" );
+CALL mtr.add_suppression("Found an entry in the 'default_roles' table with unknown authorization ID '`mysql.pxc.sst.user`@`localhost`'");


In my opinion, this error means that something is not correct with the mysql.pxc.sst.user. We need to do figure out why we get this error.

needed because you guys were not cleaning up the user correctly.
The change in wsrep_sst about revoke role is exactly for this.

venkatesh-prasad-v · 2024-11-22T05:58:56Z

scripts/wsrep_sst_clone.sh

+#              │  Joiner starts   │
+#              └────────┬─────────┘
+#              ┌────────▼─────────┐
+#              │open NetCat listnr│


Suggested change

# │open NetCat listnr│

# │Open NetCat listnr│

venkatesh-prasad-v · 2024-11-22T06:05:03Z

scripts/wsrep_sst_clone.sh

+readonly EINVAL=22
+readonly EPIPE=32
+readonly ETIMEDOUT=110
+progress_monitor=""


progress_monitor appears unused. Please check once.

venkatesh-prasad-v · 2024-11-22T06:07:58Z

scripts/wsrep_sst_clone.sh

+CLEANUP_CLONE_SSL=""
+CLONE_USER=""
+NC_PID=""
+RP_PURGED_EMERGENCY=""


Can you please add a comment what's this variable is for?

venkatesh-prasad-v · 2024-11-22T08:34:29Z

scripts/wsrep_sst_clone.sh

+
+        # we stay on hold now, waiting for the Joiner to expose the service
+        wsrep_log_info "-> WAIT for Joiner MySQL to be available nc -w 1 -i 1 $SST_HOST_STRIPPED $WSREP_SST_OPT_REMOTE_HOSTPORT"
+        while [ 1 == 1 ]; do


For better readability

Suggested change

while [ 1 == 1 ]; do

while true; do

venkatesh-prasad-v · 2024-11-22T08:52:27Z

scripts/wsrep_sst_clone.sh

+        CLONE_SSL_CA="NULL"
+    else
+        CLEANUP_CLONE_PLUGIN=""
+        CLONE_SSL_CERT=`$MYSQL_ACLIENT -e "SELECT @@clone_ssl_cert"`


In my tests, the following command failed with the following error and thus SSL was not used.

2024-11-22T05:39:30.038722Z 0 [Note] [MY-000000] [WSREP-SST] ERROR 1193 (HY000) at line 1: Unknown system variable 'clone_ssl_cert'

Is there any race condition because of which this can happen?

venkatesh-prasad-v · 2024-11-22T08:55:10Z

scripts/wsrep_sst_clone.sh

+    if [ -z "$CLIENT_SSL_CERT" -o -z "$CLIENT_SSL_KEY" ]
+    then
+        CLIENT_SSL_CERT=$(parse_cnf client ssl_cert "")
+        CLIENT_SSL_KEY=$(parse_cnf client ssl_key "")
+        CLIENT_SSL_CA=$(parse_cnf client ssl_ca "")
+        CLIENT_SSL_MODE=$(parse_cnf client ssl_mode "")
+    fi


As a fallback method, why dont we read in the mysqld section for the ssl certificates?

The XB SST script reads the certs from mysql section if it was not specified in the [sst] section

Ref: https://github.com/Tusamarco/percona-xtradb-cluster/blob/PXC-4469-sst-clone/scripts/wsrep_sst_xtrabackup-v2.sh#L722-L733

venkatesh-prasad-v · 2024-11-22T09:08:09Z

scripts/wsrep_sst_clone.sh

+        CLONE_EXECUTE_SQL=$(wsrep_mktemp_in_dir "$WSREP_SST_OPT_DATA" --suffix=.sql clone_execute_XXXX)
+        CLONE_PREPARE_SQL=$(wsrep_mktemp_in_dir "$WSREP_SST_OPT_DATA" --suffix=.sql clone_prepare_XXXX)
+
+        cleanup_donor # Remove potentially existing clone user


Here we call cleanup_donor to remove the potentially existing clone user, however this will run a DROP USER IF EXISTS 'clone_sst'@'%' in TOI and is replicated to other nodes as well. Is this something we want?

2024-11-22T08:58:32.975616Z 0 [Note] [MY-000000] [WSREP-SST] (debug) Cleanup MySQL ADMIN_PSWD: yx9!A-ezgMrQz5Q7HdKLwD7rEqHzJ0Ot 2024-11-22T08:58:32.975760Z 0 [Note] [MY-000000] [WSREP-SST] (debug) Cleanup MySQL MYSQL_ACLIENT: /home/venki/work/pxc/merge/80/bld/bin/mysql --no-defaults -umysql.pxc.sst.user -S/home/venki/work/pxc/merge/80/bld/mysql-test/var/tmp/mysqld.1.sock --batch --skip_column_names --silent 2024-11-22T08:58:32.986956Z 13 [Note] [MY-000000] [WSREP] wsrep_sync_wait: thd->variables.wsrep_sync_wait= 15, mask= 1, thd->variables.wsrep_on= 1 2024-11-22T08:58:32.987820Z 0 [Note] [MY-000000] [WSREP-SST] ERROR 1193 (HY000) at line 1: Unknown system variable 'clone_ssl_cert' 2024-11-22T08:58:32.997468Z 14 [Note] [MY-000000] [WSREP] wsrep_sync_wait: thd->variables.wsrep_sync_wait= 15, mask= 1, thd->variables.wsrep_on= 1 2024-11-22T08:58:33.004851Z 14 [Note] [MY-000000] [WSREP] Executing Query (DROP USER IF EXISTS 'clone_sst'@'%') with write-set (-1) and exec_mode: local in TO Isolation mode 2024-11-22T08:58:33.004913Z 14 [Note] [MY-000000] [WSREP] wsrep: initiating TOI for write set (-1) 2024-11-22T08:58:33.005813Z 14 [Note] [MY-000000] [WSREP] Query (DROP USER IF EXISTS 'clone_sst'@'%') with write-set (3) and exec_mode: toi replicated in TO Isolation mode 2024-11-22T08:58:33.005835Z 14 [Note] [MY-000000] [WSREP] wsrep: TO isolation initiated for write set (3) 2024-11-22T08:58:33.011278Z 14 [Note] [MY-000000] [WSREP] TO END: 3: DROP USER IF EXISTS 'clone_sst'@'%' 2024-11-22T08:58:33.011306Z 14 [Note] [MY-000000] [WSREP] wsrep: completed TOI write set (3) 2024-11-22T08:58:33.011323Z 14 [Note] [MY-000000] [WSREP] Setting WSREPXid (InnoDB): f0d030d8-a8af-11ef-927b-272a1375415c:3 2024-11-22T08:58:33.011372Z 14 [Note] [MY-000000] [WSREP] Updating WSREPXid: f0d030d8-a8af-11ef-927b-272a1375415c:3 2024-11-22T08:58:33.013101Z 14 [Note] [MY-000000] [WSREP] TO END: -1 2024-11-22T08:58:33.013812Z 14 [Warning] [MY-000000] [WSREP] Toggling wsrep_on to OFF will affect sql_log_bin. Check manual for more details 2024-11-22T08:58:33.019913Z 0 [Note] [MY-000000] [WSREP-SST] Cleanup DONOR DONE. 2024-11-22T08:58:33.019989Z 0 [Note] [MY-000000] [WSREP-SST] (debug) -> PREPARE DONOR 2024-11-22T08:58:33.033988Z 15 [Note] [MY-000000] [WSREP] wsrep_sync_wait: thd->variables.wsrep_sync_wait= 15, mask= 1, thd->variables.wsrep_on= 1 2024-11-22T08:58:33.034750Z 15 [Note] [MY-000000] [WSREP] wsrep_sync_wait: thd->variables.wsrep_sync_wait= 15, mask= 1, thd->variables.wsrep_on= 1 2024-11-22T08:58:33.040082Z 0 [Note] [MY-000000] [WSREP-SST] Installing CLONE plugin

good catch!
This is probably better to run this on the donor only!

venkatesh-prasad-v · 2024-11-22T09:15:12Z

scripts/wsrep_sst_clone.sh

+        wsrep_log_info "Server SSL settings on $ROLE: CERT=$SERVER_SSL_CERT, KEY=$SERVER_SSL_KEY, CA=$SERVER_SSL_CA"
+        wsrep_log_info "Client SSL settings on $ROLE: CERT=$CLIENT_SSL_CERT, KEY=$CLIENT_SSL_KEY, CA=$CLIENT_SSL_CA"
+        wsrep_log_info "CLONE SSL settings on $ROLE: CERT=$CLONE_SSL_CERT, KEY=$CLONE_SSL_KEY, CA=$CLONE_SSL_CA"
+        REQUIRE_SSL="REQUIRE SSL"


Here, we set the REQURIE_SSL variable to REQUIRE SSL, but it gets overwritten in line number 428? Is this expected?

This is what we will get in the error log even when both server and client have the ssl certificates in place.

2024-11-22T09:05:44.965420Z 0 [Note] [MY-000000] [WSREP-SST] Installing CLONE plugin 2024-11-22T09:05:45.086583Z 0 [Note] [MY-000000] [WSREP-SST] CLONE SSL not configured. Checking general MySQL SSL settings. 2024-11-22T09:05:45.086607Z 0 [Note] [MY-000000] [WSREP-SST] Using SSL configuration from MySQL Server. 2024-11-22T09:05:45.112218Z 0 [Note] [MY-000000] [WSREP-SST] Server SSL settings on recipient: CERT=/home/venki/work/pxc/merge/80/mysql-test/std_data/server-cert.pem, KEY=/home/venki/work/pxc/merge/80/mysql-test/std_data/server-key.pem, CA=/home/venki/work/pxc/merge/80/mysql-test/std_data/cacert.pem 2024-11-22T09:05:45.112236Z 0 [Note] [MY-000000] [WSREP-SST] Client SSL settings on recipient: CERT=/home/venki/work/pxc/merge/80/mysql-test/std_data/server-cert.pem, KEY=/home/venki/work/pxc/merge/80/mysql-test/std_data/server-key.pem, CA=/home/venki/work/pxc/merge/80/mysql-test/std_data/cacert.pem 2024-11-22T09:05:45.112256Z 0 [Note] [MY-000000] [WSREP-SST] CLONE SSL settings on recipient: CERT=/home/venki/work/pxc/merge/80/mysql-test/std_data/server-cert.pem, KEY=/home/venki/work/pxc/merge/80/mysql-test/std_data/server-key.pem, CA=/home/venki/work/pxc/merge/80/mysql-test/std_data/cacert.pem 2024-11-22T09:05:45.112321Z 0 [Note] [MY-000000] [WSREP-SST] SSL Not supported yet.

Yes I have intentionally disabled SSL until we are ok with the global process.
Then will focus on SSL. Have SSL between the Donor/joiner is not a big problem if the certificates are correctly installed. More cumbersome is to deal with the temp instance doing the clone action. But most of the logic is already in the script, I just ignore it in favour of the clone mechanics.

kamil-holubicki · 2024-11-22T08:57:48Z

mysql-test/suite/galera/t/galera_sst_clone.cnf

+[sst]
+netcat_port=4442
+wsrep-debug=true
+clone_instance_port=4444


I think we should remove these new params from cnf after rework

kamil-holubicki · 2024-11-22T08:58:34Z

scripts/wsrep_sst_clone.sh

+#              │Waits for IST            │
+#              └─────────────────────────┘
+#----------------------------------------------------------------------------
+# Config variables:


Not relevant anymore

kamil-holubicki · 2024-11-22T09:51:02Z

scripts/wsrep_sst_clone.sh

+    WSREP_SST_OPT_REMOTE_JOINER_PSWD=""
+
+    echo "continue" # donor can resume updating data
+    WSREP_SST_OPT_REMOTE_AUTH=$(echo $WSREP_SST_OPT_ADDR_LOCAL | cut -d '@' -f 1)


Right now --address parameter is in format clone_user:clone_passwd@joiner_ip:joiner:port
Why does it not conform to the format expected by --address in sst_common (ip:port/path)?
We are doing the full parsing here instead of taking the advantage of
what we already have in wsrep_sst_common.sh
If it was like joiner_ip:joiner_port/clone_user:clone_passwd
wsrep_sst_common, parses this into tokens
WSREP_SST_OPT_HOST -> joiner_ip
WSREP_SST_OPT_PORT -> joiner_port
WSREP_SST_OPT_PATH -> clone_user:clone_passwd
Then we just need to split WSREP_SST_OPT_PATH here.

kamil-holubicki · 2024-11-22T09:52:02Z

scripts/wsrep_sst_clone.sh

+    WSREP_SST_OPT_REMOTE_HOST_WITH_PORT=$(echo $WSREP_SST_OPT_ADDR_LOCAL | cut -d '@' -f 2)
+    WSREP_SST_OPT_REMOTE_HOST=$(echo $WSREP_SST_OPT_REMOTE_HOST_WITH_PORT | cut -d ':' -f 1)
+    WSREP_SST_OPT_REMOTE_HOSTPORT=$(echo $WSREP_SST_OPT_REMOTE_HOST_WITH_PORT | cut -d ':' -f 2)
+    TSST_PORT=$WSREP_SST_OPT_REMOTE_HOSTPORT


We don't need two variables for the same. Let's remove TSST_PORT and use WSREP_SST_OPT_REMOTE_HOSTPORT

kamil-holubicki · 2024-11-22T09:52:51Z

scripts/wsrep_sst_clone.sh

+            cat $CLONE_PREPARE_SQL >> /dev/stderr
+            exit $RC
+        fi
+        #Before waiting for the Joiner Clone mysql we send out the message this is a SST


Multiple occurences.
Comments format is inconsistent. Please add a space after leading hash.

kamil-holubicki · 2024-11-22T09:56:08Z

scripts/wsrep_sst_clone.sh

+        JOINER_CLONE_HOST="$WSREP_SST_OPT_ADDR_LOCAL"
+        JOINER_CLONE_PORT=$CLONE_INSTANCE_PORT
+    fi
+    TSST_PORT=$JOINER_CLONE_PORT


Two variables for the same. Let's use JOINER_CLONE_PORT (or rename it just to JOINER_SST_PORT and remove TSST_PORT

kamil-holubicki · 2024-11-22T09:57:15Z

scripts/wsrep_sst_clone.sh

+
+    # Report clone credentials/address to the caller
+    wsrep_log_debug "-> ready passing string |$CLONE_USER:$CLONE_PSWD]$JOINER_CLONE_HOST:$JOINER_CLONE_PORT|"
+    echo "ready $CLONE_USER:$CLONE_PSWD]$JOINER_CLONE_HOST:$JOINER_CLONE_PORT"


Let's just pass it as $JOINER_CLONE_HOST:$JOINER_CLONE_PORT/$CLONE_USER:$CLONE_PSWD
This way it will be a lot easier to parse it on donor side (already in sst_common). Please see other related comments.

kamil-holubicki · 2024-11-22T09:58:33Z

scripts/wsrep_sst_clone.sh

+
+    wsrep_log_debug "-> WAIT DIR DONOR MESSAGE DONE"
+
+    if [[ ! `cat $WSREP_SST_OPT_DATA/XST_FILE.txt` =~ "SST@" ]];then


We still didn't solve the problem of not full message in file
My proposal is to send from Donor such a string:
SST@${WSREP_SST_OPT_GTID}<EOF>
and then wait in a loop above until we receive <EOF> substring in a file
then we can test the content of the file.
BTW. In the case of IST, shouldn't Donor signal by sending something?

Here we have 2 messages.
SST or IST, there is NO other. So if not SST is IST, or as you fear a broken one. No need to add code to check the obvious (IST)

kamil-holubicki · 2024-11-22T09:59:59Z

scripts/wsrep_sst_clone.sh

+    wsrep_log_debug "-> WAIT DIR DONOR MESSAGE DONE"
+
+    if [[ ! `cat $WSREP_SST_OPT_DATA/XST_FILE.txt` =~ "SST@" ]];then
+        wsrep_log_info "DONOR SAY IST"


I don't see anything being sent by Donor in case of IST. We should send something like
IST<EOF>. This will allow the Joiner to immediately leave SST script.

kamil-holubicki · 2024-11-22T10:01:49Z

sql/wsrep_sst.cc

@@ -1540,7 +1542,7 @@ static bool is_sst_request_valid(const std::string &msg) {
      Instead of this we will just allow alpha-num + a few special characters
      (colon, slash, dot, underscore, square brackets, hyphen). */
    std::string data = msg.substr(method_len + 1, data_len);
-    static const std::regex allowed_chars_regex("[\\w:/.[\\]-]+");
+    static const std::regex allowed_chars_regex("[\\w:/.[\\]@-]+");


I think we don't need this change if we pass the address in the form ip:port/path

Tusamarco and others added 20 commits August 2, 2024 15:26

inserting clone scripts

8f5dfa0

Working version of clone

13e77e4

This version has a patch to go straight to ist so is not fully functional unless removing that bypass

First version that works SST and IST

876f274

Working copy of wsrep_clone with PXC

0c5887c

SSL is NOT supported

adding needed GRANTS

b721c41

adding MTR tests

d44cab3

commenting unneeded debug

93315c9

fixing debug setting

1c2799b

Update mysql_system_users.sql

b92a948

bug. Wrong comma

INFO not DEBUG

3795899

trying to identify why is not reporting the clone status in the mtr test

9c852c8

fixing config missed wsrep allowed methods

trying to identify why is not reporting the clone status in the mtr test

d7bab63

fixing config missed wsrep allowed methods

trying to identify why is not reporting the clone status in the mtr test

f5ff5cc

fixing config missed wsrep allowed methods

Working version also on MTR. I still need to find a way to be sure ne…

e307fc3

…tcat is not running

SST clone with mtr successful

b8ff4e3

pre release

a971965

fix mtr

09d9a78

fixing result

87b3971

clean up and netcat removal

caa7bb1

Adding some help description

c7fdebb

Tusamarco assigned kamil-holubicki Oct 7, 2024

kamil-holubicki requested review from venkatesh-prasad-v and kamil-holubicki October 7, 2024 17:14

Tusamarco and others added 5 commits October 18, 2024 11:39

Change needed by different NetCat implementation on Debian

ab1fa17

Add the clone script to CMake to install them in the right location.

b74a565

Need to ad extension or make script will not identify them correctly

fix in the message

f9cf9d8

adjusting value to 1 given NC RH implementation does not support 0

b28c4d7

Merge branch 'percona:8.0' into PXC-4469-sst-clone

91938a8

Tusamarco and others added 3 commits October 30, 2024 17:51

fix permission adding DELETE

5832eb8

Merge branch 'percona:8.0' into PXC-4469-sst-clone

0f1dae1

Adding revoke on SST user to remove the role association before drop …

db53e22

…the user Some edit/cleanup as suggested by Kamil

kamil-holubicki reviewed Nov 12, 2024

View reviewed changes

scripts/wsrep_sst_common_clone.sh Outdated Show resolved Hide resolved

scripts/wsrep_sst_common_clone.sh Outdated Show resolved Hide resolved

Tusamarco added 2 commits November 19, 2024 19:11

Forgot to change the Cmake list

76663b0

venkatesh-prasad-v requested changes Nov 22, 2024

View reviewed changes

kamil-holubicki requested changes Nov 22, 2024

View reviewed changes




		GRANT ALTER, CREATE, SELECT, INSERT ON PERCONA_SCHEMA.xtrabackup_history TO 'mysql.pxc.sst.role'@localhost;

		CALL mtr.add_suppression("Found an entry in the 'role_edges' table with unknown authorization ID '`mysql.pxc.sst.user`@`localhost`'" );
		CALL mtr.add_suppression("Found an entry in the 'default_roles' table with unknown authorization ID '`mysql.pxc.sst.user`@`localhost`'");


		wsrep_log_debug "-> WAIT DIR DONOR MESSAGE DONE"

		if [[ ! `cat $WSREP_SST_OPT_DATA/XST_FILE.txt` =~ "SST@" ]];then

Pxc 4469 sst clone #1961

Are you sure you want to change the base?

Pxc 4469 sst clone #1961

Conversation

Tusamarco commented Oct 7, 2024

it-percona-cla commented Oct 7, 2024 • edited Loading

Tusamarco commented Oct 7, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kamil-holubicki commented Nov 12, 2024

Tusamarco commented Nov 20, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kamil-holubicki Nov 22, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kamil-holubicki Nov 22, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

it-percona-cla commented Oct 7, 2024 •

edited

Loading

kamil-holubicki Nov 22, 2024 •

edited

Loading

kamil-holubicki Nov 22, 2024 •

edited

Loading