Missing letters in Node and Chart Id in Alerts tab as well as in the tab titles (when title is a node) in Netdata Cloud

Help
,
1

Alright guys. This is just weird. Figured I would check if anyone else has seen this before…

On the Alerts tab in Netdata Cloud, all of my node name as well as the associated ChartIDs for the alarm seem to be missing random letters. Every row in those columns are missing characters, but each seem to be in different positions in the node name.

This can also be seen in the node list in the right side filter. However, some of the node names in the list are displayed properly without any missing characters. Even though the node name is as long or longer.

The characters that are omitted will be for example, the 2nd ,5th ,11th ,13th ,16th and 18th characters on the first row. The next row however will omit the 2nd ,6th ,9th ,14th ,16th ,19th and 21st. It seems to vary per row. This happens with ALL rows. All nodes (various hostnames) as well as the ChartID associated with that alert. Sometimes the ChartID is missing the end of the chart. (For example “disk_space._” instead of “disk_space._home” “disk_space._mnt” etc etc).

It appears that the majority of the time, the missing characters seem to be vowels…

Just like with the nodes though, the ChartIDs don’t always seem to be missing characters (other than the 2nd half of the chart) however, sometimes they DO. For example “wb_lg_mtrcs.excld_rqsts” which should actually be “web_log_metrics.excluded_requests”.

For the node name, it will also omit the same characters. Example: “mtrcs” for node name instead of “metrics”

The really weird part is that on both the “Nodes” and “Anomalies” tab (and associated filters), everything looks fine. No issues with the node names or any missing characters.

It really is weird and I’ve been banding my head with it for probably a month at least and hopes that maybe one of the nightly updates would fix it but no luck. These machines have been running for years with these names and there were 0 changes made to any of the nodes, configs etc for quite awhile leading up to this. Just started happening out of the blue a couple months ago maybe… It is not intermittent.

I am unable to find anywhere in the configs or the logs where the nodes are actually referenced with the incorrect/missing characters.

This happens as/for multiple users, using Google sign-in, on different browsers (Chrome, FF).

Other spaces/rooms created by and/or with access by the same users do not exhibit the same issue. Seems to only be contained to just that one space.

All nodes are running on Debian 10/11 and installed/updated from apt and running the same versions as the parent node’s info below:

#netdata -v
netdata v1.37.0-71-nightly
#netdata -W buildinfo
Version: netdata v1.37.0-71-nightly
Configure options:  '--build=x86_64-linux-gnu' '--includedir=${prefix}/include' '--mandir=${prefix}/share/man' '--infodir=${prefix}/share/info' '--disable-silent-rules' '--libdir=${prefix}/lib/x86_64-linux-gnu' '--libexecdir=${prefix}/lib/x86_64-linux-gnu' '--disable-maintainer-mode' '--prefix=/usr' '--sysconfdir=/etc' '--localstatedir=/var' '--libdir=/usr/lib' '--libexecdir=/usr/libexec' '--with-user=netdata' '--with-math' '--with-zlib' '--with-webdir=/var/lib/netdata/www' '--disable-dependency-tracking' 'build_alias=x86_64-linux-gnu' 'CFLAGS=-g -O2 -fdebug-prefix-map=/usr/src/netdata=. -fstack-protector-strong -Wformat -Werror=format-security' 'LDFLAGS=-Wl,-z,relro' 'CPPFLAGS=-Wdate-time -D_FORTIFY_SOURCE=2' 'CXXFLAGS=-g -O2 -fdebug-prefix-map=/usr/src/netdata=. -fstack-protector-strong -Wformat -Werror=format-security'
Install type: binpkg-deb
    Binary architecture: x86_64
    Packaging distro:
Features:
    dbengine:                   YES
    Native HTTPS:               YES
    Netdata Cloud:              YES
    ACLK:                       YES
    TLS Host Verification:      YES
    Machine Learning:           YES
    Stream Compression:         NO
Libraries:
    protobuf:                YES (system)
    jemalloc:                NO
    JSON-C:                  YES
    libcap:                  NO
    libcrypto:               YES
    libm:                    YES
    tcalloc:                 NO
    zlib:                    YES
Plugins:
    apps:                    YES
    cgroup Network Tracking: YES
    CUPS:                    YES
    EBPF:                    YES
    IPMI:                    YES
    NFACCT:                  YES
    perf:                    YES
    slabinfo:                YES
    Xen:                     NO
    Xen VBD Error Tracking:  NO
Exporters:
    AWS Kinesis:             NO
    GCP PubSub:              NO
    MongoDB:                 NO
    Prometheus Remote Write: YES
Debug/Developer Features:
    Trace Allocations:       NO
#

Everything also seems ok in “netdatacli aclk-state”. I have opted not to publicly post that here but if needed, will obfuscate the IDs, node/host names etc.

Any ideas/suggestions/etc would be appreciated.

Thanks!

2

Hey @DevNull ,

Thanks for raising this concern with and sorry for causing you to bang your head around this :sweat_smile:

We have some places in the UI where due to the available display size we apply ellipsis when the element name is long. We decided to apply the ellipsis in the middle and not at the end because in some cases the suffix of a name could be the most important part. This still isn’t a straightforward thing because it is difficult to decide where/how to apply it. We decided to get some parts of the name hoping that users could still be able to identify what it is. Also in this cases on mouse hover you are able to see the full name, one example below:

image
image750×136 8.73 KB

This isn’t just see on the alert tab but probably where it is applied more often because there is a lot of information on that table. See another example on a char context name on the Nodes tab:

image
image1270×219 37.4 KB

Are the nodes having more or less the same name length on the other spaces/rooms? The only guess I have is that maybe the nodes names are smaller.

Hope this helps bring some clarity to why you are seeing this.

Regards,
Hugo

3

Yes sir. It does show properly on hover. That was also one of the reasons I was sure that nothing was broken as functionality seemed fine. I just for the life of me couldn’t figure out why. Never even occurred to me that it might be by design. :man_facepalming:t5:

Of course, now it makes PERFECT sense as the problem you are trying to mitigate is one that used to annoy me. :rofl:

Correct. That is exactly what I was seeing.

Yep. Most of them follow the same structure and some are long. What got me is that in some places it abbreviates them, like on my Alerts and the name of the tab it creates when you view the node. On the “Nodes” tab though, none were shortened so I was confused. Also on the anomalies tab and in the filter they aren’t shortened. I’m assuming because the size of those elements affords more room than the Node column in Alerts.

Thank Hugo! You sir are a LEGEND! That was fast! I’m going to have to get you a cape here soon if you keep saving me like this. :superhero:t2: lol

Thanks again!

1 Like
4

Glad to know that we at least solved it :grin:

Yes, it really depends on the size of the elements where we want to place the text,

He he, no worries I guess I just lucky to be browsing the forum.

Cheers,
Hugo