You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 

441 lines
22 KiB

diff --git a/heartbeat/SAPHana b/heartbeat/SAPHana
index 1913dc3..ed0443b 100644
--- a/heartbeat/SAPHana
+++ b/heartbeat/SAPHana
@@ -48,6 +48,8 @@ HANA_STATE_SECONDARY=1
HANA_STATE_STANDALONE=2
HANA_STATE_DEFECT=3
+debug_attributes=0
+
SH=/bin/sh
#
@@ -132,19 +134,19 @@ function saphana_meta_data() {
<?xml version="1.0"?>
<!DOCTYPE resource-agent SYSTEM "ra-api-1.dtd">
<resource-agent name="SAPHana">
-<version>0.149.4</version>
+<version>0.149.7</version>
<shortdesc lang="en">Manages two SAP HANA instances in system replication (SR).</shortdesc>
<longdesc lang="en">
The SAPHanaSR resource agent manages two SAP Hana instances (databases) which are configured
-in system replication. This first version is limitted to the scale-up scenario. Scale-Out is
+in system replication. This first version is limited to the scale-up scenario. Scale-Out is
not supported in this version.
Managing the two SAP HANA instances means that the resource agent controls the start/stop of the
instances. In addition the resource agent is able to monitor the SAP HANA databases to check their
availability on landscape host configuration level. For this monitoring the resource agent relies on interfaces
provided by SAP. A third task of the resource agent is to also check the synchronisation status
-of the two SAP HANA databases. If the synchronisation is not "SOK", than the cluster avoids to
+of the two SAP HANA databases. If the synchronisation is not "SOK", then the cluster avoids to
failover to the secondary side, if the primary fails. This is to improve the data consistency.
The resource agent uses the following four interfaces provided by SAP:
@@ -162,7 +164,7 @@ The resource agent uses the following four interfaces provided by SAP:
3. hdbnsutil
The interface hdbnsutil is used to check the "topology" of the system replication as well as the current configuration
- (primary/secondary) of a SAP HANA database instance. A second task of the interface is the posibility to run a
+ (primary/secondary) of a SAP HANA database instance. A second task of the interface is the possibility to run a
system replication takeover (sr_takeover) or to register a former primary to a newer one (sr_register).
4. hdbsql / systemReplicationStatus
@@ -198,7 +200,7 @@ The resource agent uses the following four interfaces provided by SAP:
</parameter>
<parameter name="AUTOMATED_REGISTER" unique="0" required="0">
<shortdesc lang="en">Define, if a former primary should automatically be registered.</shortdesc>
- <longdesc lang="en">The parameter AUTOMATED_REGISTER defines, wether a former primary instance should
+ <longdesc lang="en">The parameter AUTOMATED_REGISTER defines, whether a former primary instance should
be registered automatically by the resource agent during cluster/resource start, if the DUPLICATE_PRIMARY_TIMEOUT is expired... TDB
</longdesc>
<content type="boolean" default="false" />
@@ -207,7 +209,7 @@ The resource agent uses the following four interfaces provided by SAP:
<shortdesc lang="en">Time difference needed between to primary time stamps, if a dual-primary situation occurs</shortdesc>
<longdesc lang="en">Time difference needed between to primary time stamps,
if a dual-primary situation occurs. If the time difference is
- less than the time gap, than the cluster hold one or both instances in a "WAITING" status. This is to give a admin
+ less than the time gap, then the cluster hold one or both instances in a "WAITING" status. This is to give an admin
a chance to react on a failover. A failed former primary will be registered after the time difference is passed. After
this registration to the new primary all data will be overwritten by the system replication.
</longdesc>
@@ -316,7 +318,7 @@ function remoteHost2remoteNode()
# descript: is_clone : find out if we are configured to run in a Master/Slave configuration
# rc: 0: it is a clone, 1: it is not a clone
#
-# DONE: PRIO2: For the first shippment (scale-out) we need to limit the clones to 2
+# DONE: PRIO2: For the first shipment (scale-out) we need to limit the clones to 2
#
function is_clone() {
super_ocf_log info "FLOW $FUNCNAME ($*)"
@@ -356,8 +358,14 @@ function get_hana_attribute()
local attr_node=$1
local attr_name=$2
local attr_store=${3:-reboot} # DONE: PRIO5 get this (optional) from parameter
- local attr_default=${4:-}
- crm_attribute -N ${attr_node} -G -n "$attr_name" -l $attr_store -q -d "$attr_default"; rc=$?
+ local attr_default=${5:-}
+ local attr_val=""
+ attr_val=$(crm_attribute -N ${attr_node} -G -n "$attr_name" -l $attr_store -q -d "$attr_default"); rc=$?
+ if [ $debug_attributes -eq 1 ]; then
+ dstr=$(date)
+ echo "$dstr: SAPHana: crm_attribute -N ${attr_node} -G -n \"$attr_name\" -l $attr_store -q --> $attr_val" >> /var/log/fhATTRIBUTE
+ fi
+ echo "$attr_val"
super_ocf_log info "FLOW $FUNCNAME rc=$rc"
return $rc
}
@@ -381,6 +389,10 @@ function set_hana_attribute()
if [ "$attr_old" != "$attr_value" ]; then
super_ocf_log debug "DBG: SET attribute $attr_name for node ${attr_node} to ${attr_value} former ($attr_old) get_rc=$get_rc "
crm_attribute -N $attr_node -v $attr_value -n "$attr_name" -l $attr_store; rc=$?
+ if [ $debug_attributes -eq 1 ]; then
+ dstr=$(date)
+ echo "$dstr: SAPHana: crm_attribute -N $attr_node -v $attr_value -n \"$attr_name\" -l $attr_store" >> /var/log/fhATTRIBUTE
+ fi
else
super_ocf_log debug "DBG: LET attribute $attr_name for node ${attr_node} still be ${attr_value}"
rc=0
@@ -448,7 +460,7 @@ scoring_crm_master()
local roles="$1"
local sync="$2"
local skip=0
- local myScore=-1
+ local myScore=""
for scan in "${SCORING_TABLE_PREFERRED_SITE_TAKEOVER[@]}"; do
if [ $skip -eq 0 ]; then
read rolePatt syncPatt score <<< $scan
@@ -461,7 +473,10 @@ scoring_crm_master()
fi
done
super_ocf_log debug "DBG: scoring_crm_master adjust score $myScore"
- set_crm_master $myScore
+ # TODO: PRIO1: DO Not Score, If we did not found our role/sync at this moment - bsc#919925
+ if [ -n "$myScore" ]; then
+ set_crm_master $myScore
+ fi
}
#
@@ -1068,6 +1083,27 @@ function saphana_start_primary()
case "$lpa_dec" in
0 ) # LPA says start-up
lpa_advice="start"
+ # TODO: PRIO1: We need to do a special handling for remote being a 234-Secondary in SR Status SOK
+ # if ( remote_role like [234]:S ) && ( remote_sync_status is SOK|PRIM ) && ( PreferSiteTakeover )
+ # then lpa_advice="wait"
+ remoteRole=$(get_hana_attribute $remoteNode ${ATTR_NAME_HANA_ROLES[@]})
+ remoteSync=$(get_hana_attribute $remoteNode ${ATTR_NAME_HANA_SYNC_STATUS[@]})
+ super_ocf_log info "DEC: saphana_primary - checking remoteStatus"
+ if ocf_is_true "${PreferSiteTakeover}"; then
+ remoteStatus="$remoteRole:$remoteSync"
+ case "$remoteStatus" in
+ [234]:S:*:SOK | [234]:S:*:PRIM )
+ lpa_advice="wait"
+ # TODO: PRIO3: Split WAIT into WAIT4TAKEOVER
+ super_ocf_log info "DEC: saphana_primary - waiting for secondary to takeover (SOK, PreferSiteTakover)"
+ ;;
+ * )
+ super_ocf_log info "DEC: saphana_primary - remoteStatus is: $remoteStatus"
+ ;;
+ esac
+ else
+ super_ocf_log info "DEC: saphana_primary - PreferSiteTakeover set to false"
+ fi
;;
1) # LPA says register!
lpa_advice="register"
@@ -1075,7 +1111,7 @@ function saphana_start_primary()
2) # LPA says wait for second LPT
lpa_advice="wait"
;;
- 3 | 4 ) # LPA says something is completely wrong - FAIL resource
+ 3 | 4 ) # LPA says something is completely wrong - FAIL resource # TODO: PRIO1: RC3 for waiting remote side to report lss
lpa_advice="fail"
;;
* ) # LPA failed with an unkonown status - FAIL resource
@@ -1098,7 +1134,7 @@ function saphana_start_primary()
super_ocf_log info "LPA: landcape: UP, LPA: start ==> keep running"
LPTloc=$(date '+%s')
lpa_set_lpt $LPTloc
- rc=$OCF_SUCCSESS
+ rc=$OCF_SUCCESS
;;
1 ) # landcape says we are down, lets start and adjust scores and return code
super_ocf_log info "LPA: landcape: DOWN, LPA: start ==> start instance"
@@ -1149,7 +1185,7 @@ function saphana_start_primary()
case "$lss" in
2 | 3 | 4 ) # as we ARE up we just keep it up
# TODO: PRIO3: I now change from "just keep it up to take that down"
- # TODO: PRIO3: OCF_SUCCSESS, OCF_NOT_RUNNING or OCF_ERR_xxxx ?
+ # TODO: PRIO3: OCF_SUCCESS, OCF_NOT_RUNNING or OCF_ERR_xxxx ?
set_crm_master -9000
#scoring_crm_master "$my_role" "$my_sync"
rc=$OCF_ERR_GENERIC
@@ -1159,7 +1195,7 @@ function saphana_start_primary()
# TODO: PRIO3: Check, if WAITING is correct here
set_hana_attribute ${NODENAME} "WAITING" ${ATTR_NAME_HANA_CLONE_STATE[@]}
set_crm_master -9000
- rc=$OCF_SUCCSESS
+ rc=$OCF_SUCCESS
;;
esac
;;
@@ -1277,7 +1313,7 @@ function saphana_start_secondary()
super_ocf_log info "ACT: PRIMARY seams to be down now ==> WAITING"
set_hana_attribute ${NODENAME} "WAITING" ${ATTR_NAME_HANA_CLONE_STATE[@]}
set_crm_master -INFINITY
- rc=$OCF_SUCCSESS
+ rc=$OCF_SUCCESS
fi
else
lpa_set_lpt 30
@@ -1286,7 +1322,7 @@ function saphana_start_secondary()
super_ocf_log info "ACT: wait_for_primary_master ==> WAITING"
set_hana_attribute ${NODENAME} "WAITING" ${ATTR_NAME_HANA_CLONE_STATE[@]}
set_crm_master -INFINITY
- rc=$OCF_SUCCSESS
+ rc=$OCF_SUCCESS
fi
super_ocf_log info "FLOW $FUNCNAME rc=$rc"
return $rc
@@ -1453,7 +1489,8 @@ function lpa_init_lpt() {
# LPTlocal > LPTremore ===> rc=0 (start)
# LPTRemote > LPTlocal ===> rc=1 (register)
# Stalemate in all other cases ==> STALEMATE-HANDLING ===> rc=2 (wait)
-# LPTRemote is not initialized (0)
+# LPTRemote is not initialized or node not kown in cluster (crm_mon -l) (0)
+# TODO: PRIO1: Need to introduce a return-code 3 for remote sides lpa not ready
# THEN:
# WAIT ==> like STALEMATE-HANDLING ===> rc=2 (wait)
#
@@ -1625,7 +1662,6 @@ function saphana_monitor_primary()
else
super_ocf_log info "LPA: Dual primary detected and AUTOMATED_REGISTER='false' ==> WAITING"
fi
-
return $OCF_SUCCESS
fi
promoted=0;
@@ -1853,11 +1889,11 @@ function saphana_monitor_secondary()
scoring_crm_master "$my_role" "$my_sync"
;;
"SFAIL" ) # This is currently NOT a possible node to promote
- super_ocf_log info "DEC: secondary with sync status FAILED ==> EXCLUDE as posible takeover node"
+ super_ocf_log info "DEC: secondary with sync status FAILED ==> EXCLUDE as possible takeover node"
set_crm_master -INFINITY
;;
"*" ) # Unknown sync status
- super_ocf_log info "DEC: secondary with sync status UKNOWN/UNDEFINED ==> EXCLUDE as posible takeover node"
+ super_ocf_log info "DEC: secondary with sync status UKNOWN/UNDEFINED ==> EXCLUDE as possible takeover node"
set_crm_master -INFINITY
;;
esac
@@ -1889,10 +1925,12 @@ function saphana_monitor_clone() {
local rc=$OCF_ERR_GENERIC
local promoted=0
local init_attribute=0
+ local lpaRc=0
+ local mRc=0
+ local myMaster=-1
my_role=$(get_hana_attribute ${NODENAME} ${ATTR_NAME_HANA_ROLES[@]})
my_sync=$(get_hana_attribute ${NODENAME} ${ATTR_NAME_HANA_SYNC_STATUS[@]})
- lpa_check_lpt_status # TODO: PRIO3 : remove that line later - its only to call lpa_check_lpt_status much more often for checking
if ocf_is_probe; then
super_ocf_log debug "DBG: PROBE ONLY"
@@ -1904,6 +1942,16 @@ function saphana_monitor_clone() {
#
check_for_primary; primary_status=$?
if [ $primary_status -eq $HANA_STATE_PRIMARY ]; then
+ # FIX: bsc#919925 Leaving Node Maintenance stops HANA Resource Agent
+ # TODO: PRIO1: Maybe we need a lpa-check here to
+ if ocf_is_probe; then
+ myMaster=$(get_crm_master); mRc=$?
+ if [ $mRc -ne 0 ]; then
+ set_crm_master 5
+ elif [ $myMaster -eq -1 ]; then
+ set_crm_master 5
+ fi
+ fi
saphana_monitor_primary; rc=$?
else
if [ $primary_status -eq $HANA_STATE_SECONDARY ]; then
diff --git a/heartbeat/SAPHanaTopology b/heartbeat/SAPHanaTopology
index 082ad29..1d4887f 100644
--- a/heartbeat/SAPHanaTopology
+++ b/heartbeat/SAPHanaTopology
@@ -14,6 +14,7 @@
# Support: linux@sap.com
# License: GNU General Public License (GPL)
# Copyright: (c) 2014 SUSE Linux Products GmbH
+# (c) 2015 SUSE Linux GmbH
#
# An example usage:
# See usage() function below for more details...
@@ -39,6 +40,8 @@ HANA_STATE_SECONDARY=1
HANA_STATE_STANDALONE=2
HANA_STATE_DEFECT=3
+debug_attributes=0
+
SH=/bin/sh
#
@@ -123,7 +126,7 @@ function sht_meta_data() {
<?xml version="1.0"?>
<!DOCTYPE resource-agent SYSTEM "ra-api-1.dtd">
<resource-agent name="SAPHanaTopology">
- <version>0.149.4</version>
+ <version>0.149.6</version>
<shortdesc lang="en">Analyzes SAP HANA System Replication Topology.</shortdesc>
<longdesc lang="en">This RA analyzes the SAP HANA topology and "sends" all findings via the node status attributes to
all nodes in the cluster. These attributes are taken by the SAPHana RA to control the SAP Hana Databases.
@@ -205,7 +208,13 @@ function get_hana_attribute()
local attr_node=$1
local attr_name=$2
local attr_store=${3:-reboot} # DONE: PRIO5 get this (optional) from parameter
- crm_attribute -N ${attr_node} -G -n "$attr_name" -l $attr_store -q; rc=$?
+ local attr_val=""
+ attr_val=$(crm_attribute -N ${attr_node} -G -n "$attr_name" -l $attr_store -q); rc=$?
+ if [ $debug_attributes -eq 1 ]; then
+ dstr=$(date)
+ echo "$dstr: SAPHanaTopology: crm_attribute -N ${attr_node} -G -n \"$attr_name\" -l $attr_store -q --> $attr_val" >> /var/log/fhATTRIBUTE
+ fi
+ echo "$attr_val"
if [ $rc -ne 0 ]; then
super_ocf_log debug "DBG: ATTRIBUTE-FAILURE: crm_attribute -N $attr_node -G -n "$attr_name" -l $attr_store -q"
fi
@@ -230,6 +239,10 @@ function set_hana_attribute()
attr_old=$(get_hana_attribute $attr_node $attr_name $attr_store); get_rc=$?
if [ "$attr_old" != "$attr_value" ]; then
super_ocf_log debug "DBG: SET attribute $attr_name for node ${attr_node} to ${attr_value} former ($attr_old) get_rc=$get_rc "
+ if [ $debug_attributes -eq 1 ]; then
+ dstr=$(date)
+ echo "$dstr: SAPHanaTopology: crm_attribute -N $attr_node -v $attr_value -n \"$attr_name\" -l $attr_store" >> /var/log/fhATTRIBUTE
+ fi
crm_attribute -N $attr_node -v "$attr_value" -n "$attr_name" -l $attr_store; rc=$?
if [ $rc -ne 0 ]; then
super_ocf_log debug "DBG: ATTRIBUTE-FAILURE: crm_attribute -N $attr_node -v $attr_value -n "$attr_name" -l $attr_store"
@@ -377,18 +390,32 @@ function sht_init() {
*openais* ) nodelist=$(crm_node -l | awk '/member/ {print $2}');;
*cman* ) nodelist=$(crm_node -l);;
esac
+ #### SAP-CALL
hdbANSWER=$(su - ${sidadm} -c "hdbnsutil -sr_state --sapcontrol=1" 2>/dev/null)
super_ocf_log debug "DBG2: hdbANSWER=\$\(su - ${sidadm} -c \"hdbnsutil -sr_state --sapcontrol=1\"\)"
site=$(echo "$hdbANSWER" | awk -F= '/site name/ {print $2}')
srmode=$(echo "$hdbANSWER" | awk -F= '/mode/ {print $2}')
- MAPPING=$(echo "$hdbANSWER" | awk -F[=/] '$1 ~ "mapping" && $3 !~ site { print $4 }' site=$site)
+ if [ $debug_attributes -eq 1 ]; then
+ dstr=$(date)
+ echo "$dstr: SAPHanaTopology: srmode=$srmode" >> /var/log/fhATTRIBUTE
+ fi
+ MAPPING=$(echo "$hdbANSWER" | awk -F[=/] '$1 == "mapping" && $3 != site { print $4 }' site=$site)
super_ocf_log debug "DBG: site=$site, mode=$srmode, MAPPING=$MAPPING"
#
# filter all non-cluster mappings
#
- hanaRemoteHost=$(for n1 in $nodelist; do for n2 in $MAPPING; do if [ "$n1" == "$n2" ]; then echo $n1; fi; done; done )
- super_ocf_log info "DEC: site=$site, mode=$srmode, MAPPING=$MAPPING, hanaRemoteHost=$hanaRemoteHost"
- super_ocf_log debug "DBG: site=$site, mode=$srmode, MAPPING=$MAPPING, hanaRemoteHost=$hanaRemoteHost"
+ # DONE: PRIO2: Need mapping between HANA HOSTS not cluster NODES
+ local hanaVHost
+ hanaRemoteHost=$(for n1 in $nodelist; do
+ hanaVHost=$(get_hana_attribute ${n1} ${ATTR_NAME_HANA_VHOST[@]})
+ for n2 in $MAPPING; do
+ if [ "$hanaVHost" == "$n2" ]; then
+ echo $hanaVHost;
+ fi;
+ done;
+ done )
+ super_ocf_log info "DEC: site=$site, mode=$srmode, MAPPING=$MAPPING, hanaRemoteHost=$hanaRemoteHost"
+ super_ocf_log debug "DBG: site=$site, mode=$srmode, MAPPING=$MAPPING, hanaRemoteHost=$hanaRemoteHost"
super_ocf_log info "FLOW $FUNCNAME rc=$OCF_SUCCESS"
return $OCF_SUCCESS
}
@@ -422,6 +449,7 @@ function check_for_primary() {
super_ocf_log err "ACT: check_for_primary: we didn't expect node_status to be: <$node_status>"
dump=$( echo $node_status | hexdump -C );
super_ocf_log err "ACT: check_for_primary: we didn't expect node_status to be: DUMP <$dump>"
+ #### SAP-CALL
node_full_status=$(su - ${sidadm} -c "hdbnsutil -sr_state" 2>/dev/null )
node_status=$(echo "$node_full_status" | awk '$1=="mode:" {print $2}')
super_ocf_log info "DEC: check_for_primary: loop=$i: node_status=$node_status"
@@ -440,6 +468,7 @@ function check_for_primary() {
#
function start_saphostagent()
{
+ ### SAP-CALL
if [ -x "${HOSTEXEC_PATH}" ]; then
${HOSTEXEC_PATH} pf=${HOSTEXEC_PROFILE_PATH}
fi
@@ -453,9 +482,10 @@ function start_saphostagent()
#
function stop_saphostagent()
{
- if [ -x "${HOSTEXEC_PATH}" ]; then
- ${HOSTEXEC_PATH} -stop
- fi
+ ### SAP-CALL
+ if [ -x "${HOSTEXEC_PATH}" ]; then
+ ${HOSTEXEC_PATH} -stop
+ fi
}
#
@@ -586,7 +616,7 @@ function sht_validate() {
#
function sht_start_clone() {
super_ocf_log info "FLOW $FUNCNAME ($*)"
- local rc=$OCF_NOT_RUNNING
+ local rc=$OCF_NOT_RUNNING
sht_start; rc=$?
return $rc
}
@@ -666,27 +696,30 @@ function sht_monitor_clone() {
# DONE: PRIO1: ASK: Is the output format of ListInstances fix? Could we take that as an API?
# try to catch: Inst Info : LNX - 42 - lv9041 - 740, patch 36, changelist 1444691
# We rely on the following format: SID is word#4, NR is work#6, vHost is word#8
+ #### SAP-CALL
vName=$(/usr/sap/hostctrl/exe/saphostctrl -function ListInstances \
| awk '$4 == SID && $6=NR { print $8 }' SID=$SID NR=$InstanceNr 2>/dev/null )
- super_ocf_log debug "DBG: ListInstances: $(/usr/sap/hostctrl/exe/saphostctrl -function ListInstances)"
+ # super_ocf_log debug "DBG: ListInstances: $(/usr/sap/hostctrl/exe/saphostctrl -function ListInstances)"
if [ -n "$vName" ]; then
set_hana_attribute ${NODENAME} "$vName" ${ATTR_NAME_HANA_VHOST[@]}
else
vName=$(get_hana_attribute ${NODENAME} ${ATTR_NAME_HANA_VHOST[@]})
fi
#site=$(get_site_name)
+ #### SAP-CALL
hanaANSWER=$(su - $sidadm -c "python exe/python_support/landscapeHostConfiguration.py" 2>/dev/null); hanalrc="$?"
hanarole=$(echo "$hanaANSWER" | tr -d ' ' | awk -F'|' '$2 == host { printf "%s:%s:%s:%s\n",$10,$11,$12,$13 } ' host=${vName})
#if [ -z "$MAPPING" ]; then
# super_ocf_log info "ACT: Did not find remote Host at this moment"
#fi
- # FH TODO PRIO1: TRY TO GET RID OF "ATTR_NAME_HANA_REMOTEHOST"
+ # FH TODO PRIO3: TRY TO GET RID OF "ATTR_NAME_HANA_REMOTEHOST"
if [ -n "$hanaRemoteHost" ]; then
set_hana_attribute ${NODENAME} "$hanaRemoteHost" ${ATTR_NAME_HANA_REMOTEHOST[@]}
fi
set_hana_attribute ${NODENAME} "$hanalrc:$hanaPrim:$hanarole" ${ATTR_NAME_HANA_ROLES[@]}
- set_hana_attribute ${NODENAME} "$site" ${ATTR_NAME_HANA_SITE[@]}
- set_hana_attribute ${NODENAME} "$vName" ${ATTR_NAME_HANA_VHOST[@]}
+ if [ -n "$site" ]; then
+ set_hana_attribute ${NODENAME} "$site" ${ATTR_NAME_HANA_SITE[@]}
+ fi
case "$hanaPrim" in
P ) ;;
S ) # only secondary may propargate its sync status
@@ -701,7 +734,6 @@ function sht_monitor_clone() {
done
;;
esac
- #ATTR_NAME_HANA_STATUS # TODO: PRIO5: For SCALE-OUT: Fill that attribute later
super_ocf_log info "FLOW $FUNCNAME rc=$rc"
return $rc
}