no space left on device -- but there is

1,446 views
Skip to first unread message

Barry Kaplan

unread,
Jul 5, 2016, 1:12:08 PM7/5/16
to Consul
On one of our nodes we see:

2016/07/05 17:09:46 [ERR] serf: Failed to update snapshot: write /var/lib/consul/serf/local.snapshot: no space left on device

but:

$ df
Filesystem      Size  Used Avail Use% Mounted on
udev            
7.9G   12K  7.9G   1% /dev
tmpfs           1.6G  620K  1.6G   1% /
run
/dev/xvda1       16G  4.5G   11G  30% /
none            4.0K     0  4.0K   0% /
sys/fs/cgroup
none            
5.0M     0  5.0M   0% /run/lock
none            
7.9G  260K  7.9G   1% /run/shm
none            
100M     0  100M   0% /run/user
/dev/xvdf        99G  1.1G   93G   2% /data

What device could it be talking about?

Brian Lalor

unread,
Jul 5, 2016, 1:14:08 PM7/5/16
to Consul
Check your inode usage.  `df -i`.

--
This mailing list is governed under the HashiCorp Community Guidelines - https://www.hashicorp.com/community-guidelines.html. Behavior in violation of those guidelines may result in your removal from this mailing list.
 
GitHub Issues: https://github.com/hashicorp/consul/issues
IRC: #consul on Freenode
---
You received this message because you are subscribed to the Google Groups "Consul" group.
To unsubscribe from this group and stop receiving emails from it, send an email to consul-tool...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/consul-tool/3646d703-94f9-458b-916c-c074f7541928%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

— 
Brian Lalor

Barry Kaplan

unread,
Jul 5, 2016, 1:14:41 PM7/5/16
to Consul
I restarted the client and the error stopped. I'm clueless.

Michael Fischer

unread,
Jul 5, 2016, 3:39:21 PM7/5/16
to consu...@googlegroups.com
Which version of Consul are you using?  Prior to 0.6 it would create a 16 or 32 GB memory mapped file and if your data filesystem didn't have enough space it would cause this error. 

The solution is to ensure you have at least that much free space or to upgrade to 0.6. 

On Tuesday, July 5, 2016, Barry Kaplan <mem...@gmail.com> wrote:
I restarted the client and the error stopped. I'm clueless.

--
This mailing list is governed under the HashiCorp Community Guidelines - https://www.hashicorp.com/community-guidelines.html. Behavior in violation of those guidelines may result in your removal from this mailing list.
 
GitHub Issues: https://github.com/hashicorp/consul/issues
IRC: #consul on Freenode
---
You received this message because you are subscribed to the Google Groups "Consul" group.
To unsubscribe from this group and stop receiving emails from it, send an email to consul-tool...@googlegroups.com.

Barry Kaplan

unread,
Jul 6, 2016, 11:44:34 PM7/6/16
to Consul
This is version 0.6.3

Barry Kaplan

unread,
Aug 21, 2016, 9:33:02 PM8/21/16
to Consul
Just encountered this problem again with 0.6.3, on dozens of nodes. No evidence of disk shortage or inode shortage. I'm quite certain this is not a consul issue, but I cannot find the cause.


Failed to update snapshot: write /var/lib/consul/serf/local.snapshot: no space left on device

[s-mesos-slave-5 10.0.136.120]$ df
Filesystem      Size  Used Avail Use% Mounted on
udev            
7.9G   12K  7.9G   1%
/dev
tmpfs           1.6G  968K  1.6G   1% /
run
/dev/xvda1       16G  8.4G  6.6G  56% /

none            4.0K     0  4.0K   0% /
sys/fs/cgroup
none            
5.0M     0  5.0M   0% /run/lock
none            
7.9G  2.2M  7.9G   1% /run/shm
none            
100M     0  100M   0% /run/user
/dev/xvdf        99G  4.6G   89G   5% /data
bindfs           16G  8.4G  6.6G  56% /
log

[s-mesos-slave-5 10.0.136.120]$ df -i
Filesystem     Inodes IUsed IFree IUse% Mounted on
udev            
2.0M   391  2.0M    1% /dev
tmpfs            2.0M   36K  2.0M    2% /
run
/dev/xvda1       1.0M  198K  827K   20% /
none             2.0M    11  2.0M    1% /
sys/fs/cgroup
none            
2.0M     1  2.0M    1% /run/lock
none            
2.0M    43  2.0M    1% /run/shm
none            
2.0M     3  2.0M    1% /run/user
/dev/xvdf        6.3M  208K  6.1M    4% /data
bindfs           1.0M  198K  827K   20% /
log


Jonathan Ballet

unread,
Aug 22, 2016, 5:20:41 AM8/22/16
to consu...@googlegroups.com

I faced this issue some weeks ago, due to a full disk at some point, but even after making some space, Consul couldn't recover.

There's an issue opened for this with a proposed fix, but I haven't tested yet, see https://github.com/hashicorp/consul/issues/1744

--
This mailing list is governed under the HashiCorp Community Guidelines - https://www.hashicorp.com/community-guidelines.html. Behavior in violation of those guidelines may result in your removal from this mailing list.
 
GitHub Issues: https://github.com/hashicorp/consul/issues
IRC: #consul on Freenode
---
You received this message because you are subscribed to the Google Groups "Consul" group.
To unsubscribe from this group and stop receiving emails from it, send an email to consul-tool...@googlegroups.com.

Barry Kaplan

unread,
Aug 24, 2016, 4:26:23 AM8/24/16
to Consul
Hmm, that could have been the case for me. Thanks.
Reply all
Reply to author
Forward
0 new messages