Hello list! We have been attempting to set up a MySQL Cluster using 4 Solaris 10 x86 servers. Two managers + mysqld, and two data nodes.
Initially, we used versions: SunOS mysql01.unix 5.11 snv_40 i86pc i386 i86pc SunOS mysql02.unix 5.11 snv_40 i86pc i386 i86pc SunOS mysqldata01.unix 5.11 snv_40 i86pc i386 i86pc SunOS mysqldata02.unix 5.11 snv_40 i86pc i386 i86pc Supermicro 6014L-T, 2GB memory, 2x dual 2Ghz XEONs. http://www.supermicro.com/products/system/1U/6014/SYS-6014L-T.cfm MySQL is configured to use: NoOfReplicas=2, DataMemory=1200M, IndexMemory=200M. Although MySQL 5.0 runs just fine, as it uses only memory, we ran out far too quickly. So we attempted 5.1.22rc But during our stress test, no more than 5 minutes into it, one, or more often that not, both; mysqldata servers would hang. (Only the data nodes). One such test is: CREATE LOGFILE GROUP lg_1 ADD UNDOFILE 'undo_1.dat' INITIAL_SIZE 16M UNDO_BUFFER_SIZE 2M ENGINE NDB; Symptoms of hang is.. well, a total hang, but it would ping. You can telnet to port 22 and it would connect, but you would not get a greeting. Console has no response. ssh session sitting in "top" would simply just stop. (but not disconnect). Disk BUSY led is permanently on, so it is very very busy doing something on the disks. 2 disks setup in mirror. But I also tried removing mirror, and running on single drive with same results. I also ran a "top > output.txt" but that too just hung. I have yet to manage to get any useful data out of it hanging. Power cycle brings it back, but nothing in logs, nor /var/crash/. Since then I have tried: Latest gcc. No difference. Latest SunPro CC. No difference. SunOS mysqldata01.unix 5.11 snv_60b i86pc i386 i86pc SunOS mysqldata02.unix 5.11 snv_60b i86pc i386 i86pc No difference. SunOS mysqldata01.unix 5.11 snv_70b i86pc i386 i86pc SunOS mysqldata02.unix 5.11 snv_70b i86pc i386 i86pc No difference. Right now, I am considering jumping Solaris 9 and testing if that will fix the situation. Is there anything anyone can suggest we try also? Both in terms of fixing the crashes, but also in terms of monitoring the server in an attempt to solve why it hangs. Any reply is appreciated! Lund -- Jorgen Lundman | <[EMAIL PROTECTED]> Unix Administrator | +81 (0)3 -5456-2687 ext 1017 (work) Shibuya-ku, Tokyo | +81 (0)90-5578-8500 (cell) Japan | +81 (0)3 -3375-1767 (home) _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
