by Vikram Khatri | Mar 16, 2017
I have seen working with hundreds of customers a chronic LOG FULL condition and the shortcut people use is to increase the LOGPRIMARY and LOGSECOND. This solution works well for a single unit of work that can not fit into the total log space available. So, first...
by Vikram Khatri | Mar 10, 2017
In Linux environment, it is a common practice to upgrade the kernel for security patches, vulnerability fixes etc. And if you are running a DB2 pureScale environment, it is likely that you are going to break the GPFS GPL as soon as you upgrade the kernel and this will...
by Vikram Khatri | Mar 10, 2017
During pureScale instance creation, you might run into instance creation failed condition. Under some circumstances, it is possible that db2icrt does not perform a proper clean-up for GPFS and it may fail next time complaining that the disk is already a GPFS disk....
by Vikram Khatri | Mar 7, 2017
The SQLs while DB2 is up and running can be captured in different ways such as: Query column STMT_TEXT from table functions such as MON_GET_ACTIVITY or WLM_GET_WORKLOAD_OCCURRENCE_ACTIVITIES, and MON_GET_ACTIVITY_DETAILS –> SQLs that have been submitted but...
by Vikram Khatri | Feb 23, 2017
Here are two templates that one can use to make Automatic Client Reroute to work properly in a HADR environment. Please note that these instructions are only valid for HADR pair in which the second machine is ready to takeover. I will publish / update this article...
by Vikram Khatri | Jan 5, 2017
This is about how to rebuild TSA resources in HADR environment if you have used db2haicu in past but added resources such as tie-breaker disks that are not supported by the db2haicu tool. So, in case of a difficulty – how do you get back soon. You can always...
by Vikram Khatri | Sep 29, 2016
When you get errors like this: “SQL1517N db2start failed because the cluster manager resource states are inconsistent.” or something like “Cluster manager resource states for the DB2 instance are inconsistent. Refer to the db2diag.log for more...
by Vikram Khatri | Jul 1, 2016
RSCT License Issue $ db2start 128 SQL1677N DB2START or DB2STOP processing failed due to a DB2 cluster services error DATA #9 : SQLHA Remote Command Output, PD_TYPE_SQLHA_COMMAND_RESPONSE, 3508 bytes commandResponse->callRC: 0x00000000 commandResponse->output:...
by Vikram Khatri | Jun 27, 2016
GPL Compile Problem When you try to install DB2 pureScale on Linux platform (RHEL or SLES), you may receive the error. ERROR: A missing Linux package is preventing the compilation of the IBM General Parallel File System (GPFS) Portability Layer (GPL) module. The...
by Vikram Khatri | Jun 1, 2016
If you try to add a member to a running cluster, you may get the following db2start error. ERROR: The “db2start add MEMBER 2 hostname node04.purescale.ibm.local netname node04.purescale.ibm.local port 0 -iupdt” command failed with the return code:...
by Vikram Khatri | May 28, 2016
One of my most favorite icons on GNOME has been Dropline Neu! for many years and I still have a copy of it since 2006. It is now hard to find. Download ICON-DroplineNeu.tar.bz2 And, copy to /tmp directory # cd /usr/share/icons # tar xvfj /tmp/ICON-DroplineNeu.tar.bz2...
by Vikram Khatri | May 20, 2016
If you need to drop a member in DB2 pureScale, the db2iupdt process requires that member to be available during the drop process. But, the actual world is more cruel than what db2iupdt thinks. For example: If you have a 3 member cluster and someone moved / migrated...
by Vikram Khatri | May 19, 2016
If you install DB2 pureScale in Linux, the Fast I/O Fencing does not get enabled automatically. If you refer to Knowledge Center for DB2, you will find number of steps to enable fast I/O fencing. After DB2 instance is created and databases are created, I usually run...
by Vikram Khatri | Apr 30, 2016
Use dmidecode -t slot to find out the slot numbers that are in USE and the Bus Address. Use that bus address in lspci -s <busaddress> to find out what is in the slot. For example: I have NIC in slot-7, Q-Logic in Slot – 2 and Mellanox in Slot – 3. I...
by Vikram Khatri | Apr 30, 2016
A FC card has a World Wide Name (WWN). Card may have two ports and each port is known as by its World Wide Port Name (WWPN). I seemed to have confused myself with WWN, WWNN and WWPN and didn’t understand the difference as this is not my topic. In the example...
by Vikram Khatri | Apr 24, 2016
In order to be a successful DB2 pureScale DBA, one has to learn how to diagnose and fix TSA. Though, this is not necessary but it will help to fix cluster in case you run into situation when you get messages like “SQL1517N db2start failed because the cluster...
by Vikram Khatri | Apr 21, 2016
This is for for my own reference and does not constitute any recommendations. VIO Server Changes Verify flow_control is set to yes – This is the default setting lsattr –El ent0 | grep flow and lsattr –El ent1 | grep flow chdev -l ent0 –a flow_control=yes chdev...
by Vikram Khatri | Apr 20, 2016
It is possible to avoid tie breaker disk in pureScale and HADR just by following a simple majority rule. You need a tie breaker disk in pureScale and HADR if number of hosts are even. So, in HADR you have only two nodes so your need a tie breaker disk. But, even in...
by Vikram Khatri | Apr 20, 2016
Reset RSCT Configuration (completely remove all cluster configuration) # /usr/sbin/rsct/install/bin/recfgct Change the tie-breaker # chrsrc -c IBM.PeerNode OpQuorumTieBreaker=’Operator’ Find out which tie-breaker is in use # lsrsrc -c IBM.PeerNode...
by Vikram Khatri | Apr 20, 2016
Read reservations: /usr/bin/sg_persist –in –no-inquiry –read-reservation –device=/dev/sdc /usr/bin/sg_persist –in –no-inquiry –read-key –device=/dev/sdc Prepare for cancel: /usr/bin/sg_persist –out...
