cluster-commits February 2011

cluster-commits@lists.fedorahosted.org

7 participants
135 discussions

cluster: RHEL57 - resource-agents: Remove netmask from IP address when creating list of them
by Marek Grác 25 Feb '11

25 Feb '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=56… Commit: 56b618c99947d457579c6bc9f0a1de4aae329595 Parent: f26a3627a1087816fae1bb671623fc74bf653a90 Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Mon Feb 7 14:14:39 2011 +0100 Committer: Marek 'marx' Grac <mgrac(a)redhat.com> CommitterDate: Wed Feb 23 18:39:55 2011 +0100 resource-agents: Remove netmask from IP address when creating list of them IP address were copied directly to configuration of various resource agents. Resolves: rhbz#675739 --- rgmanager/src/resources/utils/config-utils.sh | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/rgmanager/src/resources/utils/config-utils.sh b/rgmanager/src/resources/utils/config-utils.sh index 5646fa8..c54b266 100644 --- a/rgmanager/src/resources/utils/config-utils.sh +++ b/rgmanager/src/resources/utils/config-utils.sh @@ -210,7 +210,9 @@ build_ip_list() break fi - ipaddrs="$ipaddrs $ipaddr" + # remove netmask + iponly=`echo $ipaddr | sed 's/\/.*//'` + ipaddrs="$ipaddrs $iponly" ((x++)) done

1 0

cluster: RHEL57 - resource-agents: Add option disable_rdisc to ip.sh
by Marek Grác 25 Feb '11

25 Feb '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=f2… Commit: f26a3627a1087816fae1bb671623fc74bf653a90 Parent: b9306e19ba3f346326befd933c8ecdd9bc8973bf Author: Marek 'marx' Grac <mgrac(a)redhat.com> AuthorDate: Fri Jan 28 15:29:48 2011 +0100 Committer: Marek 'marx' Grac <mgrac(a)redhat.com> CommitterDate: Wed Feb 23 18:38:08 2011 +0100 resource-agents: Add option disable_rdisc to ip.sh rdisc is called by ip.sh which causes static routes to be updated by dynamic ones. This option adds possibility to disable this feature. Resolves: rhbz#620700 --- rgmanager/src/resources/ip.sh | 21 +++++++++++++++++++-- 1 files changed, 19 insertions(+), 2 deletions(-) diff --git a/rgmanager/src/resources/ip.sh b/rgmanager/src/resources/ip.sh index 4f7e432..38d1ab9 100755 --- a/rgmanager/src/resources/ip.sh +++ b/rgmanager/src/resources/ip.sh @@ -121,6 +121,17 @@ meta_data() <content type="string"/> </parameter> + <parameter name="disable_rdisc"> + <longdesc lang="en"> + Disable updating of routing using RDISC protocol and + preserve static routes. + </longdesc> + <shortdesc lang="en"> + Disable updating of routing using RDISC protocol + </shortdesc> + <content type="boolean"/> + </parameter> + </parameters> <actions> @@ -632,7 +643,10 @@ ipv6() # Not sure if this is necessary for ipv6 either. file=$(which rdisc 2>/dev/null) if [ -f "$file" ]; then - killall -HUP rdisc || rdisc -fs + if [ "$OCF_RESKEY_disable_rdisc" != "yes" ] && \ + [ "$OCF_RESKEY_disable_rdisc" != "1" ]; then + killall -HUP rdisc || rdisc -fs + fi fi return 0 @@ -711,7 +725,10 @@ ipv4() file=$(which rdisc 2>/dev/null) if [ -f "$file" ]; then - killall -HUP rdisc || rdisc -fs + if [ "$OCF_RESKEY_disable_rdisc" != "yes" ] && \ + [ "$OCF_RESKEY_disable_rdisc" != "1" ]; then + killall -HUP rdisc || rdisc -fs + fi fi return 0

1 0

cluster: STABLE31 - config: stop leaking memory on config reload
by Fabio M. Di Nitto 24 Feb '11

24 Feb '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=7d… Commit: 7d2b388cfe9cdc7e0eac65218788527a3969fe8c Parent: 0901934fdcd49f85deac905da3d63497d3885382 Author: Fabio M. Di Nitto <fdinitto(a)redhat.com> AuthorDate: Thu Feb 24 15:04:31 2011 +0100 Committer: Fabio M. Di Nitto <fdinitto(a)redhat.com> CommitterDate: Thu Feb 24 15:04:31 2011 +0100 config: stop leaking memory on config reload Resolves: rhbz#680155 Signed-off-by: Fabio M. Di Nitto <fdinitto(a)redhat.com> --- config/plugins/xml/config.c | 2 -- 1 files changed, 0 insertions(+), 2 deletions(-) diff --git a/config/plugins/xml/config.c b/config/plugins/xml/config.c index 1db23da..7dcf8ba 100644 --- a/config/plugins/xml/config.c +++ b/config/plugins/xml/config.c @@ -142,7 +142,5 @@ fail: if (doc) xmlFreeDoc(doc); - xmlCleanupParser(); - return err; }

1 0

gfs2-utils: the annotated tag 3.1.1 has been created
by Bob Peterson 23 Feb '11

23 Feb '11

Gitweb: http://git.fedorahosted.org/git/gfs2-utils.git?p=gfs2-utils.git;a=commitdif… Commit: ffda3929ec66f086f91adc69682285127693f137 Parent: 0000000000000000000000000000000000000000 Author: Bob Peterson <rpeterso(a)redhat.com> AuthorDate: 2011-02-23 15:23 +0000 Committer: Bob Peterson <rpeterso(a)redhat.com> CommitterDate: 2011-02-23 15:23 +0000 annotated tag: 3.1.1 has been created at ffda3929ec66f086f91adc69682285127693f137 (tag) tagging 4a727917ff403b6f386c992d1e071e3f1a795893 (commit) replaces 3.1.0 New tag for updated source tree Benjamin Marzinski (1): gfs2_grow: fix growing on full filesystems Bob Peterson (14): mkfs.gfs2 should support discard request generation GFS2: mkfs.gfs2 segfaults with 18.55TB and -b512 gfs2-utils: minor corrections to README.build fsck.gfs2: reports master/root dinodes as unused and fixes the bitmap fsck.gfs2: can't repair rgrps resulting from gfs_grow->gfs2_convert gfs2_edit: handle corrupt file systems better gfs2_edit: print large block numbers better gfs2_edit: has problems printing gfs1 journals gfs2_edit: add -d option for printing journal details gfs2_edit: Fix error message on blockalloc when outside bitmap gfs2_edit: fix careless compiler warning gfs2_edit: fix segfault in set_bitmap when block is in rgrp fsck.gfs2: segfault in pass1b gfs2_edit savemeta doesn't save all leaf blocks for large dirs David Teigland (1): gfs_controld: remove oom_adj Steven Whitehouse (5): tune: Clean up and make closer to tune2fs strings: Clean up strings libgfs2: Remove calls to gettext from libgfs2 libgfs2: Remove unused function get_sysfs_uinit() libgfs2: Move gfs2_query into gfs2_convert

1 0

gfs2-utils: master - gfs2_edit savemeta doesn't save all leaf blocks for large dirs
by Bob Peterson 23 Feb '11

23 Feb '11

Gitweb: http://git.fedorahosted.org/git/gfs2-utils.git?p=gfs2-utils.git;a=commitdif… Commit: 4a727917ff403b6f386c992d1e071e3f1a795893 Parent: 4e79f08f455038bdfb962f941676b7d5d8c21ef8 Author: Bob Peterson <rpeterso(a)redhat.com> AuthorDate: Wed Feb 23 07:13:31 2011 -0600 Committer: Bob Peterson <rpeterso(a)redhat.com> CommitterDate: Wed Feb 23 07:13:31 2011 -0600 gfs2_edit savemeta doesn't save all leaf blocks for large dirs This patch changes the gfs2_edit "savemeta" option so that it processes exhash directories the same way that fsck.gfs2 does, namely, it does a read of the hash table and saves all the leaf blocks from that. Before it was trying to traverse the blocks, but that only works for medium-size exhash directories. rhbz#679565 --- gfs2/edit/savemeta.c | 19 +++++++++++++++---- 1 files changed, 15 insertions(+), 4 deletions(-) diff --git a/gfs2/edit/savemeta.c b/gfs2/edit/savemeta.c index 666a452..844da8c 100644 --- a/gfs2/edit/savemeta.c +++ b/gfs2/edit/savemeta.c @@ -398,10 +398,21 @@ static void save_inode_data(int out_fd) } } /* Process directory exhash inodes */ - if (S_ISDIR(inode->i_di.di_mode)) { - if (inode->i_di.di_flags & GFS2_DIF_EXHASH) { - save_indirect_blocks(out_fd, cur_list, metabh, - height, 0); + if (S_ISDIR(inode->i_di.di_mode) && + inode->i_di.di_flags & GFS2_DIF_EXHASH) { + uint64_t leaf_no, old_leaf = -1; + int li; + + for (li = 0; li < (1 << inode->i_di.di_depth); li++) { + gfs2_get_leaf_nr(inode, li, &leaf_no); + if (leaf_no == old_leaf || + gfs2_check_range(&sbd, leaf_no) != 0) + continue; + old_leaf = leaf_no; + mybh = bread(&sbd, leaf_no); + if (gfs2_check_meta(mybh, GFS2_METATYPE_LF) == 0) + save_block(sbd.device_fd, out_fd, leaf_no); + brelse(mybh); } } if (inode->i_di.di_eattr) { /* if this inode has extended attributes */

1 0

cluster: RHEL57 - fsck.gfs2: segfault in pass1b
by Bob Peterson 22 Feb '11

22 Feb '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=b9… Commit: b9306e19ba3f346326befd933c8ecdd9bc8973bf Parent: 415fc5f19db6780c76dfc00f680a9617dacd7b31 Author: Bob Peterson <rpeterso(a)redhat.com> AuthorDate: Tue Feb 22 16:40:37 2011 -0600 Committer: Bob Peterson <rpeterso(a)redhat.com> CommitterDate: Tue Feb 22 16:40:37 2011 -0600 fsck.gfs2: segfault in pass1b The problem occurred when there were duplicate block references in a dinode but all references in the duplicate list are eventually deleted due to other corruption. The fix is an additional check whether the list is empty. rhbz#679076 --- gfs2/fsck/pass1b.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/gfs2/fsck/pass1b.c b/gfs2/fsck/pass1b.c index a0c5c88..9abc8c1 100644 --- a/gfs2/fsck/pass1b.c +++ b/gfs2/fsck/pass1b.c @@ -558,7 +558,7 @@ static int handle_dup_blk(struct gfs2_sbd *sbp, struct duptree *b) last_reference = clear_a_reference(sbp, b, &b->ref_inode_list, &dh, 0); - if (last_reference) { + if (last_reference && !osi_list_empty(&b->ref_inode_list)) { uint8_t q; /* If we're down to a single reference (and not all references

1 0

fence: master - fenced: don't ignore victim_done messages for reduced victims
by David Teigland 22 Feb '11

22 Feb '11

Gitweb: http://git.fedorahosted.org/git/fence.git?p=fence.git;a=commitdiff;h=73c0f5… Commit: 73c0f58f071d713bd3b634ffcfb1d02feba84fcc Parent: 7181608a3a497bff3d9d1306e77503dbfc4c32f5 Author: David Teigland <teigland(a)redhat.com> AuthorDate: Tue Feb 22 15:47:24 2011 -0600 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Tue Feb 22 16:40:30 2011 -0600 fenced: don't ignore victim_done messages for reduced victims When a victim is "reduced" (i.e. fenced skips fencing it because it rejoins the cluster cleanly before fenced fences it), it is immediately removed from the list of victims, before the "victim_done" message is sent for it. The victim_done message updates the time of the last successful fencing operation for a failed node. The code that processes received victim_done messages was ignoring the message for the reduced victim because the node couldn't be found in the victims list. This caused the latest fencing information to not be recorded for the node, causing dlm_controld to wait indefinately for fencing to complete for the reduced victim. The fix is to simply record the information from a victim_done message even if the node is not in the victims list. bz 678704 Acked-by: Ryan O'Hara <rohara(a)redhat.com> Signed-off-by: David Teigland <teigland(a)redhat.com> --- fence/fenced/cpg.c | 18 ++++++++++++------ 1 files changed, 12 insertions(+), 6 deletions(-) diff --git a/fence/fenced/cpg.c b/fence/fenced/cpg.c index eec2ba6..716d4c4 100644 --- a/fence/fenced/cpg.c +++ b/fence/fenced/cpg.c @@ -654,9 +654,9 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) node = get_node_victim(fd, id->nodeid); if (!node) { + /* see comment below about no node */ log_debug("receive_victim_done %d:%u no victim nodeid %d", hd->nodeid, seq, id->nodeid); - return; } log_debug("receive_victim_done %d:%u remove victim %d time %llu how %d", @@ -672,9 +672,11 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) if (hd->nodeid == our_nodeid) { /* sanity check, I don't think this should happen; see comment in fence_victims() */ - if (!node->local_victim_done) - log_error("expect local_victim_done"); - node->local_victim_done = 0; + if (node) { + if (!node->local_victim_done) + log_error("expect local_victim_done"); + node->local_victim_done = 0; + } } else { /* save details of fencing operation from master, which master saves at the time it completes it */ @@ -682,8 +684,12 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) id->fence_how, id->fence_time); } - list_del(&node->list); - free(node); + /* we can have no node when reduce_victims() removes it, bz 678704 */ + + if (node) { + list_del(&node->list); + free(node); + } } /* we know that the quorum value here is consistent with the cpg events

1 0

cluster: STABLE31 - fenced: don't ignore victim_done messages for reduced victims
by David Teigland 22 Feb '11

22 Feb '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=09… Commit: 0901934fdcd49f85deac905da3d63497d3885382 Parent: c217c686b16f9aa8be3b2b49d017b11bf05b26fc Author: David Teigland <teigland(a)redhat.com> AuthorDate: Tue Feb 22 15:47:24 2011 -0600 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Tue Feb 22 16:39:53 2011 -0600 fenced: don't ignore victim_done messages for reduced victims When a victim is "reduced" (i.e. fenced skips fencing it because it rejoins the cluster cleanly before fenced fences it), it is immediately removed from the list of victims, before the "victim_done" message is sent for it. The victim_done message updates the time of the last successful fencing operation for a failed node. The code that processes received victim_done messages was ignoring the message for the reduced victim because the node couldn't be found in the victims list. This caused the latest fencing information to not be recorded for the node, causing dlm_controld to wait indefinately for fencing to complete for the reduced victim. The fix is to simply record the information from a victim_done message even if the node is not in the victims list. bz 678704 Acked-by: Ryan O'Hara <rohara(a)redhat.com> Signed-off-by: David Teigland <teigland(a)redhat.com> --- fence/fenced/cpg.c | 18 ++++++++++++------ 1 files changed, 12 insertions(+), 6 deletions(-) diff --git a/fence/fenced/cpg.c b/fence/fenced/cpg.c index a8629b9..99e16a0 100644 --- a/fence/fenced/cpg.c +++ b/fence/fenced/cpg.c @@ -652,9 +652,9 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) node = get_node_victim(fd, id->nodeid); if (!node) { + /* see comment below about no node */ log_debug("receive_victim_done %d:%u no victim nodeid %d", hd->nodeid, seq, id->nodeid); - return; } log_debug("receive_victim_done %d:%u remove victim %d time %llu how %d", @@ -670,9 +670,11 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) if (hd->nodeid == our_nodeid) { /* sanity check, I don't think this should happen; see comment in fence_victims() */ - if (!node->local_victim_done) - log_error("expect local_victim_done"); - node->local_victim_done = 0; + if (node) { + if (!node->local_victim_done) + log_error("expect local_victim_done"); + node->local_victim_done = 0; + } } else { /* save details of fencing operation from master, which master saves at the time it completes it */ @@ -680,8 +682,12 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) id->fence_how, id->fence_time); } - list_del(&node->list); - free(node); + /* we can have no node when reduce_victims() removes it, bz 678704 */ + + if (node) { + list_del(&node->list); + free(node); + } } /* we know that the quorum value here is consistent with the cpg events

1 0

cluster: RHEL6 - fenced: don't ignore victim_done messages for reduced victims
by David Teigland 22 Feb '11

22 Feb '11

Gitweb: http://git.fedorahosted.org/git/cluster.git?p=cluster.git;a=commitdiff;h=4e… Commit: 4eb20ad9f8e62d0739c7897d79bc1bccb94e5250 Parent: 8fd8d3f3c52b9efc09d911bd8fdcab3e8742e9cb Author: David Teigland <teigland(a)redhat.com> AuthorDate: Tue Feb 22 15:47:24 2011 -0600 Committer: David Teigland <teigland(a)redhat.com> CommitterDate: Tue Feb 22 16:33:11 2011 -0600 fenced: don't ignore victim_done messages for reduced victims When a victim is "reduced" (i.e. fenced skips fencing it because it rejoins the cluster cleanly before fenced fences it), it is immediately removed from the list of victims, before the "victim_done" message is sent for it. The victim_done message updates the time of the last successful fencing operation for a failed node. The code that processes received victim_done messages was ignoring the message for the reduced victim because the node couldn't be found in the victims list. This caused the latest fencing information to not be recorded for the node, causing dlm_controld to wait indefinately for fencing to complete for the reduced victim. The fix is to simply record the information from a victim_done message even if the node is not in the victims list. bz 678704 Acked-by: Ryan O'Hara <rohara(a)redhat.com> Signed-off-by: David Teigland <teigland(a)redhat.com> --- fence/fenced/cpg.c | 18 ++++++++++++------ 1 files changed, 12 insertions(+), 6 deletions(-) diff --git a/fence/fenced/cpg.c b/fence/fenced/cpg.c index a8629b9..99e16a0 100644 --- a/fence/fenced/cpg.c +++ b/fence/fenced/cpg.c @@ -652,9 +652,9 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) node = get_node_victim(fd, id->nodeid); if (!node) { + /* see comment below about no node */ log_debug("receive_victim_done %d:%u no victim nodeid %d", hd->nodeid, seq, id->nodeid); - return; } log_debug("receive_victim_done %d:%u remove victim %d time %llu how %d", @@ -670,9 +670,11 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) if (hd->nodeid == our_nodeid) { /* sanity check, I don't think this should happen; see comment in fence_victims() */ - if (!node->local_victim_done) - log_error("expect local_victim_done"); - node->local_victim_done = 0; + if (node) { + if (!node->local_victim_done) + log_error("expect local_victim_done"); + node->local_victim_done = 0; + } } else { /* save details of fencing operation from master, which master saves at the time it completes it */ @@ -680,8 +682,12 @@ static void receive_victim_done(struct fd *fd, struct fd_header *hd, int len) id->fence_how, id->fence_time); } - list_del(&node->list); - free(node); + /* we can have no node when reduce_victims() removes it, bz 678704 */ + + if (node) { + list_del(&node->list); + free(node); + } } /* we know that the quorum value here is consistent with the cpg events

1 0

gfs2-utils: master - fsck.gfs2: segfault in pass1b
by Bob Peterson 22 Feb '11

22 Feb '11

Gitweb: http://git.fedorahosted.org/git/gfs2-utils.git?p=gfs2-utils.git;a=commitdif… Commit: 4e79f08f455038bdfb962f941676b7d5d8c21ef8 Parent: 897437313adc8801b38ea42abc4fd07bd486ce3e Author: Bob Peterson <rpeterso(a)redhat.com> AuthorDate: Tue Feb 22 15:56:39 2011 -0600 Committer: Bob Peterson <rpeterso(a)redhat.com> CommitterDate: Tue Feb 22 15:56:39 2011 -0600 fsck.gfs2: segfault in pass1b The problem occurred when there were duplicate block references in a dinode but all references in the duplicate list are eventually deleted due to other corruption. The fix is an additional check whether the list is empty. rhbz#679076 --- gfs2/fsck/pass1b.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/gfs2/fsck/pass1b.c b/gfs2/fsck/pass1b.c index 090b412..5bbda7c 100644 --- a/gfs2/fsck/pass1b.c +++ b/gfs2/fsck/pass1b.c @@ -542,7 +542,7 @@ static int handle_dup_blk(struct gfs2_sbd *sbp, struct duptree *b) last_reference = clear_a_reference(sbp, b, &b->ref_inode_list, &dh, 0); - if (last_reference) { + if (last_reference && !osi_list_empty(&b->ref_inode_list)) { uint8_t q; /* If we're down to a single reference (and not all references

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

cluster-commits February 2011