FAQ

Search Discussions

52 discussions - 186 posts

  • This isn't good.... root:~# /usr/local/mysql-5.1/bin/mysql -S /var/lib/mysql-cluster/mysqld.sock mydb < dump.sql ERROR 1297 (HY000) at line 79: Got temporary error 4023 'Transaction aborted due to ...
    Kevin BurtonKevin Burton
    Jan 20, 2007 at 5:43 am
    Feb 1, 2007 at 12:52 pm
  • There was a thread on the list a while back about the performance of a NDB cluster restore but I can't seem to find it. I did a mysqldump of our current MyISAM database and I did a mysql foo < ...
    Kevin BurtonKevin Burton
    Jan 20, 2007 at 7:53 am
    Jan 22, 2007 at 5:16 am
  • Hi, I'm trying to import databases to the new Cluster I've just installed (5.0.27), but when on one table, I have this problem : CREATE TABLE `mdl_enrol_paypal` ( `id` int( 10 ) unsigned NOT NULL ...
    Mathieu CronMathieu Cron
    Jan 9, 2007 at 1:30 pm
    Jan 10, 2007 at 2:28 pm
  • Is there an easy way to find out the current size of the data in our cluster? We have allocated DataSize to be 400MB (which is about 75% more than what ndb_size.pl reported for our tables.) I'm ...
    MatthewMatthew
    Jan 15, 2007 at 11:00 pm
    Jan 16, 2007 at 6:43 pm
  • We're thinking about moving to a system where some records values are set to null. These would be VARCHAR(255) and so could theoretically save significant memory. Will this memory be returned to the ...
    Kevin BurtonKevin Burton
    Jan 5, 2007 at 3:20 am
    Jan 8, 2007 at 6:31 am
  • Hi guys, I have 2 quick questions which may or may not seem dense :) 1. We currently have a MySQL cluster with 3 servers setup as a management server and 2 servers each setup as storage and SQL ...
    Andrew HutchingsAndrew Hutchings
    Jan 20, 2007 at 9:32 am
    Feb 14, 2007 at 3:16 pm
  • OK........ I did the migration to 5.1.14 but I can't get "all dump 1000" to work and basically I'm flyin blind without it! In the mgm console I executed: all clusterlog statistics=15 and all dump ...
    Kevin BurtonKevin Burton
    Jan 20, 2007 at 5:40 am
    Jan 22, 2007 at 12:42 am
  • What's up with this? mysql DELETE FROM FOO; ERROR 1205 (HY000): Lock wait timeout exceeded; try restarting transaction Is this expected behavior or a bug? I shouldn't have to rewrite SQL like this to ...
    Kevin BurtonKevin Burton
    Jan 25, 2007 at 6:41 am
    Feb 14, 2007 at 6:41 pm
  • Another question... does adding more replicas increase performance (or decrease performance)? -- Founder/CEO Tailrank.com Location: San Francisco, CA AIM/YIM: sfburtonator Skype: burtonator Blog: ...
    Kevin BurtonKevin Burton
    Jan 20, 2007 at 10:18 pm
    Jan 22, 2007 at 12:58 am
  • Hello All, Newbie Cluster-er here. :) I know there are special procedures for altering your NDB tables in any way, but cannot find any definite instructions/HOWTO on doing this. We currently have a 2 ...
    MatthewMatthew
    Jan 15, 2007 at 8:11 pm
    Jan 16, 2007 at 2:57 am
  • I must be overlooking but can anyone point me to the config parameters to bind ndbd and mysql to a specific (incoming and outgoing) ip addresses. I want to make sure that all MySQL cluster traffic ...
    Kris BuytaertKris Buytaert
    Jan 15, 2007 at 1:58 pm
    Jan 15, 2007 at 10:23 pm
  • With partitioning and so forth it seems logical that NDB would use MyISAM or INNODB for its disk storage. Why does it use its own proprietary format? For example I could use NDB with MyISAM and use ...
    Kevin BurtonKevin Burton
    Jan 11, 2007 at 9:21 pm
    Jan 12, 2007 at 10:33 am
  • Hi group, Am I reading this error correctly? I can't make schema changes while one node is restarting? Is this planned to change before the final release? The cluster cant be inaccessible for eg. ...
    BrianBrian
    Jan 26, 2007 at 7:53 pm
    Jan 27, 2007 at 12:12 am
  • Hi everyone, I'm wondering how crashed data nodes are detected in the cluster? This is in relation to bug #25710 and MySQL 5.1.14. Our beta testing of the cluster is at a standstill due to this ...
    BrianBrian
    Jan 23, 2007 at 4:16 pm
    Jan 23, 2007 at 10:33 pm
  • Can someone explain why disk based cluster tables require datamemory to be increased as the number of disk based tables increase? I thought it was just indexmemory that needed to increase for storing ...
    BrianBrian
    Jan 22, 2007 at 5:34 pm
    Jan 22, 2007 at 7:32 pm
  • How bad is this upgrade going to be? There's a schema change in the 'mysql' database right? Should I just bring up a totally new instance and to a restore into the new instance? Kevin -- Founder/CEO ...
    Kevin BurtonKevin Burton
    Jan 17, 2007 at 3:01 am
    Jan 17, 2007 at 7:29 am
  • I have a data node that's not starting..... it's in perpetual boot stage..... all dump 1000 shows that it's at 0% memory and index usage while all other notes are well higher than that value. One of ...
    Kevin BurtonKevin Burton
    Jan 29, 2007 at 10:08 am
    Jan 29, 2007 at 10:28 pm
  • We're currently using bulk inserts with MyISAM to increase performance. Would INSERTs be faster on NDB if I just did them across 50 threads instead of batching them up? Seems like they would ...
    Kevin BurtonKevin Burton
    Jan 20, 2007 at 10:30 pm
    Jan 26, 2007 at 9:39 am
  • Why is CREATE TABLE so slow with NDB? Same issue with TRUNCATE TABLE..... I just did a basic create table and it took 3 minutes..... Kevin -- Founder/CEO Tailrank.com Location: San Francisco, CA ...
    Kevin BurtonKevin Burton
    Jan 21, 2007 at 2:03 am
    Jan 22, 2007 at 12:47 am
  • hi, Normally my cluster environment is working fine but when i propt the command like show variables: It gives following error: SHOW VARIABLES MySQL said: Documentation #1566 - Cannot proceed because ...
    NilnandanNilnandan
    Jan 5, 2007 at 2:19 pm
    Jan 20, 2007 at 2:10 pm
  • Hi, The MySQL Cluster docs say that to restore from a cluster backup I have to run ndb_restore once for each node in the cluster to restore it data. In my cluster this doesn't seem to be necessary, ...
    Mike KruckenbergMike Kruckenberg
    Jan 13, 2007 at 3:05 pm
    Jan 15, 2007 at 7:28 am
  • After shutting down and and restarting my cluster (MySQL 5.0) yesterday, logging stopped working. I checked that `clusterlog info` showed that logging was on and all the categories I wanted (the ...
    Eric WilliamsEric Williams
    Jan 9, 2007 at 8:05 pm
    Jan 15, 2007 at 2:55 am
  • I`ve been playing with all dump 1000 and all clusterlog statistics=15 lateley and I noticed that when i put all clusterlog statistics=0 I don't get %index and % memory usage from ndb . As soon as I ...
    Kris BuytaertKris Buytaert
    Jan 5, 2007 at 12:34 pm
    Jan 6, 2007 at 9:08 am
  • Hi all, I've seen references to this topic (and I'm sure many of you have done this before) but I can't seem to find anything concrete on how to get the following to work : - 2 servers each with 2 ...
    Yong LeeYong Lee
    Jan 2, 2007 at 7:31 am
    Jan 2, 2007 at 10:08 pm
  • Am I reading the 5.1 cluster backup and restore instructions correctly, where all the backup files created on each of the nodes must be available to restore the cluster? With 2 replicas (4 ndb nodes, ...
    BrianBrian
    Jan 29, 2007 at 5:22 pm
    Jan 30, 2007 at 2:43 am
  • Hm..... I just got this error message while trying to do a restore: ERROR 1297 (HY000) at line 104: Got temporary error 410 'REDO log files overloaded, consult online manual (decrease ...
    Kevin BurtonKevin Burton
    Jan 21, 2007 at 4:11 am
    Jan 22, 2007 at 12:50 am
  • Hi, I have created cluster environment on 4 pc (THIS IS THE TESTING ENVIRONMENT) 1-SQL node (512 RAM) 2-MGM node (512 RAM) 3-data node-1 (2GB RAM) 4-datanode-2 (2GB RAM) my config.ini is like this: ...
    NilnandanNilnandan
    Jan 3, 2007 at 2:24 pm
    Jan 4, 2007 at 9:56 pm
  • Hi again.... during our test sessions, we've found a big performance problem in our app. We have 2 tables: CREATE TABLE `isinmarkets` ( `isin` varchar(12) NOT NULL DEFAULT '', `market` varchar(8) NOT ...
    Michael HeckMichael Heck
    Jan 4, 2007 at 12:31 pm
    Jan 4, 2007 at 1:37 pm
  • hi, I have installed MySQL Cluster 5.1.14 for testing...I want to know about disk space storage in cluster. these are my questions: Q.1 can we store data in RAM or in DISK whenever we want, in ...
    NilnandanNilnandan
    Jan 1, 2007 at 9:17 am
    Jan 1, 2007 at 5:10 pm
  • Based on a thread earlier this week, if one were to deploy NDB in a highly multithreaded environment, normal errors from NDB would rise into application code which would need to be handled. I assume ...
    Kevin BurtonKevin Burton
    Jan 24, 2007 at 1:08 am
    Feb 14, 2007 at 5:55 am
  • For some reason ever since migrating to 5.1.14 the ndbd processes on all our nodes are terminating. StopOnError is set to 0 so the parent ndbd should restart the child (since it's basically ...
    Kevin BurtonKevin Burton
    Jan 25, 2007 at 8:29 pm
    Feb 3, 2007 at 4:46 am
  • Hi, I have some questions about MYSQL cluster, for redundancy and IO whats the suggested number of MYSQL hosts? Has anyone got any IO numbers? And when is disk clustering coming out? If anyone can ...
    MorganMorgan
    Jan 24, 2007 at 1:35 am
    Jan 30, 2007 at 5:25 am
  • OK... I'm about to send off a larger email about this as information on the Redo log is sort of missing from the manual. Anyway.... I assume the startup time for a data node in a cluster that's ...
    Kevin BurtonKevin Burton
    Jan 27, 2007 at 2:39 am
    Jan 28, 2007 at 2:51 pm
  • What's up with this? Mysql ALTER TABLE FOO - ADD .................... - ; ERROR 1297 (HY000): Got temporary error 1221 'REDO buffers overloaded, consult online manual (increase RedoBuffer)' from ...
    Kevin BurtonKevin Burton
    Jan 25, 2007 at 11:25 pm
    Jan 26, 2007 at 8:33 am
  • Hi Everyone, I was just wondering if it is possible to get binary packages of 5.0.32 from anywhere? I can't find them in the software downloads page of a MySQL Enterprise site, nor on the "free" ...
    Alex DaviesAlex Davies
    Jan 22, 2007 at 7:49 pm
    Jan 23, 2007 at 1:24 am
  • I can understand the reasoning. A bit confusing without the explaination though. Thanks for the clarity here. Brian Now you can have a huge leap forward in email: get the new Yahoo! Mail.
    BrianBrian
    Jan 22, 2007 at 7:33 pm
    Jan 22, 2007 at 7:36 pm
  • I'm doing a rolling restart and upgrading to 5.1.14 when I notice this error. It happened twice and I think the ndbd nodes were both masters at the time. After about 2-3 minutes the ndbd process ...
    Kevin BurtonKevin Burton
    Jan 20, 2007 at 4:27 am
    Jan 20, 2007 at 5:16 pm
  • Here's my problem: We're REALLY write bound and replication won't scale for us anymore since we're bound to the performance of an individual machine. We could buy more expensive servers for our slave ...
    Kevin BurtonKevin Burton
    Jan 18, 2007 at 10:36 pm
    Jan 20, 2007 at 12:15 am
  • Hello, Yesterday got quite strange problem with ndb_mgmd. MySQL Cluster 5.0.26 is running on 64 bits CentOS4. Cluster consists of two servers each running one copy of management node, data node and ...
    MindaugasMindaugas
    Jan 19, 2007 at 11:46 am
    Jan 20, 2007 at 12:12 am
  • Hi Cluster folk, Here's a quick question to find out how much performance decrease one will get by using 5.1's disk based storage -------- Original Message -------- Subject: Performance mysql Cluster ...
    Colin CharlesColin Charles
    Jan 18, 2007 at 2:07 am
    Jan 19, 2007 at 3:49 pm
  • Hi, I want to know that which is the main configuration parameter which are affected to the performance of cluster. I have installed MySQL Cluster 5.1.14. But i dont get enough performance. which is ...
    NilnandanNilnandan
    Jan 11, 2007 at 10:10 am
    Jan 11, 2007 at 3:55 pm
  • hi, I have MySQL CLuster 5.1.14 environment with 4 nodes. (1-MGM, 1-SQLD, 1-DN1, 1-DN2) when transaction was running the error was occured like error 2303: 'System error, node killed during node ...
    NilnandanNilnandan
    Jan 5, 2007 at 2:24 pm
    Jan 9, 2007 at 11:17 pm
  • Hi there, The MySQL community downloads for RPM systems seem to not include MySQL Cluster servers. Either there are no MySQL-Max packages (RHEL, SuSE) or there are only NDB Engine storage and tools ...
    FernandoFernando
    Jan 8, 2007 at 1:59 pm
    Jan 9, 2007 at 11:16 pm
  • Happy new year... we have a problem with mysql-cluster, which seems to be very mysterious. Perhaps, somebody here can give us a hint... What we want to do: An application wants to rebuild an ...
    Michael HeckMichael Heck
    Jan 2, 2007 at 9:14 am
    Jan 2, 2007 at 9:24 am
  • OK.... I'm *still* getting lock wait timeouts...... There's *no* IO on any of my disks except he REDO log and of course the table space (this is going to be an on disk data). I'm using mysqldump to ...
    Kevin BurtonKevin Burton
    Jan 27, 2007 at 5:09 am
    Jan 27, 2007 at 5:09 am
  • A few random questions. I noted before that there's very little documentation on behalf of how the REDO log operates. From what I can gather there are two disk logs used by NDB. The REDO log stores ...
    Kevin BurtonKevin Burton
    Jan 27, 2007 at 2:47 am
    Jan 27, 2007 at 2:47 am
  • What's the cluster roadmap look like for 5.2? Anyone? Bueler? Bueler? -- Founder/CEO Tailrank.com Location: San Francisco, CA AIM/YIM: sfburtonator Skype: burtonator Blog: feedblog.org Cell: ...
    Kevin BurtonKevin Burton
    Jan 27, 2007 at 2:25 am
    Jan 27, 2007 at 2:25 am
  • About six months I wrote the lbpool driver to handle load balancing MySQL directly in code (instead of hardware load balancers).. http://code.tailrank.com/lbpool Basically you can get really good ...
    Kevin BurtonKevin Burton
    Jan 20, 2007 at 9:27 pm
    Jan 20, 2007 at 9:27 pm
  • Hi everyone, I'm not sure how to calculate the correct value for MaxNoOfConcurrentOperations and MaxNoOfLocalOperations. I understand what the manual says, but it doesn't correspond to what I'm ...
    BrianBrian
    Jan 18, 2007 at 10:53 pm
    Jan 18, 2007 at 10:53 pm
  • We are testing pulling down the switch to which the cluster is connected. (generating a huge split brain) And we were hoping to have an automatic restart by defining StopOnError=false Theoretically ...
    Kris BuytaertKris Buytaert
    Jan 16, 2007 at 4:31 pm
    Jan 16, 2007 at 4:31 pm
Group Navigation
period‹ prev | Jan 2007 | next ›
Group Overview
groupcluster @
categoriesmysql
discussions52
posts186
users41
websitemysql.com
irc#mysql

41 users for January 2007

Kevin Burton: 44 posts Jon Stephens: 16 posts Stewart Smith: 16 posts Brian: 11 posts Mikael Ronström: 10 posts Adam Dixon: 8 posts Geert Vanderkelen: 6 posts Hartmut Holzgraefe: 6 posts Matthew Boehm: 6 posts Nilnandan: 6 posts Eric Williams: 4 posts Jonas Oreland: 4 posts Mathieu Cron: 4 posts Michael Heck: 4 posts Pekka: 3 posts Andrew Hutchings: 3 posts Kris Buytaert: 3 posts Matthew Montgomery: 3 posts Anatoly Pidruchny: 2 posts Kris Buytaert: 2 posts
show more
Archives