I have gitlab omnibus install on a linux debian server. Yesterday, the installation was upgraded to 14.12 (from 14.11) - and now I'm unable to access the web UI. On investigation, it seems that the issue was caused by a conflict of versions of ruby and passenger. I'm using an external nginx on the server, as the tail server has other virtual hosts. I have the latest version of passenger installed - 6.0.9. Gitlab latest version comes with ruby version 2.7.2 - but passenger 6.0.9 has no support for that version, it support 2.7.3 instead.
I downgraded passenger to version 6.0.8 - now the passenger_native error is resolved (by direct downloading exact version). However I'm getting the following error in the log:
Could not spawn process for application /opt/gitlab/embedded/service/gitlab-rails:
A timeout occurred while starting a preloader process.
There's not much more info. There is one question I found on SO from years ago - this one - with the same error - but it's not solved. I'm in no position to install/run docker on the server.
How can I get gitlab UI to show?
Note: this is cross-posted to ServerFault here.
Related
Trying to upgrade an existing React app running on AWS Amplify from node v16 to v18
The image used is the provided Amazon Linux 2
Expected it to install node 18 on the machine it spun up and build the application, since this is the LTS version
Here is the error I get in the build process:
2022-11-15T19:19:02.924Z [WARNING]: node: /lib64/libm.so.6: version GLIBC_2.27' not found (required by node)
node: /lib64/libc.so.6: version GLIBC_2.28' not found (required by node)
The same problem happened to me, I fixed it by adding a custom image in the build config.
You can see the steps to add a custom image in the following link
The image that worked for me was the following:
public.ecr.aws/t7o4u3y2/node-18.2.0:latest
Amazon Linux 2 includes Glibc 2.26 or under. To use a higher glibc version, you need to consider other AMI, for example Amazon Linux 2022 (al2022-ami) as it includes glibc 2.34.
You can also revert back to Node.js 16 as it uses a lower version of Glibc.
Thanks for opening this issue. Amazon Linux 2 doesn't support the right GLIBC version at the moment, we're coordinating internally to get that updated. As #saharl-neosec mentions, you can provide a custom docker image until we fully support it:
https://github.com/aws-amplify/amplify-hosting/issues/3109
I'm trying to do IBM Bluemix essential course.
As per the course instructions, I installed Eclipse (I initially got latest Eclipse version Neon instead of Mars but tried both Mars and Neon as of now for installing NodeEclipse .17, 1.0.1 or 1.0.2 or Enide.p2f file using both File import and Eclipse marketplace for installing Enide.p2f Node eclipse plugin in Eclipse).
One of the instruction (after installing IBM Bluemix plugin for eclipse) says, I need to install "Enide.p2f Eclipse NodeJS plugin" https://marketplace.eclipse.org/content/enidep2f-eclipse-nodejs-ide
I tried Eclipse Marketplace to install this. Searched "node eclipse" and selected the top listing entry for "Nodeeclipse 1.0.2 Nodeeclipse Core & Node.js (nodeeclipse-1)" or even "Enide.p2f for Nodeeclipse 1.0.1" version and clicked installed.
A couple of times, if failed during the resolution process to list the plugin components and Eclipse error log showed that the update site / repository server was not available.
Another method was to install the .p2f file directly File > Import > Install > Install from a file and selected the downloaded .p2f file which I got from here: https://raw.github.com/Nodeclipse/eclipse-node-ide/master/ENodeIDE.p2f
or
http://marketplace.eclipse.org/content/enide-eclipse-nodejs-ide/click
Using the import > install via file method, I'm getting "Unhandled event loop exception" error message as per Eclipse Error logs and it didn't install anything.
Using the market place method, I'm getting the following error mesg (per the screen shot):
For Nodeeclipse 1.0.1
https://i.imgsafe.org/1b1bf4f0fb.jpg
https://i.imgsafe.org/1b1be1c00a.jpg
https://i.imgsafe.org/1b1bc75cfc.jpg
For Nodeeclipse 1.0.2.
https://i.imgsafe.org/1afbf9bf56.jpg
https://i.imgsafe.org/1afa90e4f1.jpg
https://i.imgsafe.org/1af372340c.jpg
Error message:
HTTP Server Unknown HTTP Response Code (499):http://nodeeclipse.org/updates/compositeContext.xml
HttpComponents connection error response code 499.
OR
it shows the Enide.p2f .17, 1.0.1 or 1.0.2 (Whichever you select from the list) to install the plugin:
The following solutions are not available: Nodeclipse 1.0.2
(id=markdown.editor.feature.feature.group,
org.nodeclipse.pluginslist.feature.feature.group,
gitaddon.feature.feature.group,
de.bastiankrol.startexplorer.feature.feature.group,
com.eclipsesource.jshint.feature.feature.group,
org.nodeclipse.feature.group,
org.nodeclipse.enide.editors.jade.feature.feature.group,
org.chromium.sdk.feature.group, net.mihai-nita.ansicon.feature.group,
org.chromium.debug.feature.group,
pm.eclipse.editbox.feature.feature.group,
site=http://www.nodeclipse.org/updates/)
With Eclipse Mars, I'm getting the same behavior. More info on this can be found here as well: https://www.ibm.com/developerworks/community/forums/html/topic?id=76a96fa4-1086-41a7-94a6-4326f79dfacb&ps=#repliesPg=0
Did anybody else face this similar issue and what should i do to resolve it.
"https://marketplace.eclipse.org/content/enidep2f-eclipse-nodejs-ide"looks like an update site so I tried the Help->Install New Software... I then entered the address as a new update site but it failed as well...
I am not sure if it is the right place to ask this question. I have been trying to install clearcase client 8015 on a new machine with RHEL5u8 image. Tried googling a lot but could not resolve the issue as per suggestions on other forums.
It gets installed successfully. (As per the prompt)
listing of the views is done properly. (cleartool lsview)
When I do cleartool mount -all,
cleartool: Error: The MVFS file system is not installed or not loaded,
or the 'viewroot' is not mounted or is inaccessible: not a ClearCase
object.
logging in as root then starting clearcase, executing "/opt/rational/clearcase/etc/clearcase start"
OUTPUT:-
ClearCase is stopped Starting ClearCase ClearCase daemons: albd_server
FATAL: Error inserting mvfs
(/lib/modules/2.6.18-308.4.1.0.1.el5/kernel/fs/mvfs/mvfs.ko): Unknown
symbol in module, or unknown parameter (see dmesg) Loading the MVFS
was unsuccessful(8:1) /opt/rational/clearcase/etc/clearcase: Loading
MVFS failed
---- Please review /var/log/messages for information on the problem.
---- You may need to rebuild your kernel, rebuild your mvfs module,
---- and reboot your system. Information on how to do this can be found in
---- your Release Notes and file /var/adm/rational/clearcase/mvfs/mvfs_src/README.txt.
I have followed all the steps in /var/adm/rational/clearcase/mvfs/mvfs_src/README.txt but still cannot achieve anything.
Uninstallation also fails as:-
./linux_8015_uninstall
ERROR: Installation data has incompatible version 1.8.0; expected
1.7.0. Newer version of the Installation Manager was used on this system.
java.io.IOException: Installation data has incompatible version 1.8.0;
expected 1.7.0. Newer version of the Installation Manager was used on
this system.
00:01.33 ERROR [main]
com.ibm.cic.agent.core.application.HeadlessApplication run
Installation data has incompatible version 1.8.0; expected 1.7.0.
Newer version of the Installation Manager was used on this system.
java.io.IOException: Installation data has incompatible version 1.8.0;
expected 1.7.0. Newer version of the Installation Manager was used on
this system. java.io.IOException: Installation data has incompatible
version 1.8.0; expected 1.7.0. Newer version of the Installation
Manager was used on this system.
at com.ibm.cic.agent.internal.core.InstallRegistryParser.parse(InstallRegistryParser.java:160)
at com.ibm.cic.agent.internal.core.InstallRegistry.load(InstallRegistry.java:679)
at com.ibm.cic.agent.internal.core.InstallRegistry.openFile(InstallRegistry.java:485)
at com.ibm.cic.agent.internal.core.InstallRegistry.open(InstallRegistry.java:429)
Cannot proceed with installation, cannot uninstall either to try another version of clearcase, tried googling for the version mismatch of IBM installation manager but could not find anything addressing the issue properly for Linux.
Any suggestions on how to proceed? I am new to clearcase tool.
Usually, when I see only snapshot views are supported, and MVFS (for dynamic views) hasn't installed properly, I uninstall everything and try the all process from the start again.
In your case, that means installing or downgrading the IBM IM Installation Manager to the right version (similar to this issue).
It can be an IM update issue, as seen in this technote.
Further to #VonC's comment re: incompatible IM install,
review the following doc: MVFS not loading after kernel upgrade or is missing after installation on Linux.
I encountered a similar problem after an incompatible kernel path / ClearCase version. In my case, the error was similar to what is documented here in the post MVFS not installed - RHEL6 and I had to edit the kernel source to the same as the patch and rebuild. Then things stared up fine.
Review also the link How to manually restore a pre-built MVFS on Linux.
If the corresponding kerenl level is not provided it will not start and you will have to patch and make your own.
I recently realized that our servers were vulnerable to CVE-2014-0224 by doing an SSL Labs scan.
We are running a node.js server on our site. I read that the vulnerability was eliminated in node.js version 0.10.29, so I used nvm to install version 0.11.13 on our ubuntu 12 server. However, after restarting the node.js server, the scan still shows we're vulnerable.
Thinking that perhaps node.js was relying on the system installation of openssl rather than using their own, I upgraded that, as well. openssl version -b now shows that we have the version that was built on 1 Oct 2014 ... which should be plenty good by all accounts (upgrading to this version was how I eliminated the vulnerability from another one of our servers that was running a ruby server ... also on ubuntu 12).
However, I haven't been able to find the right combination to eliminate the vulnerabilty from our node.js server. How can I do that? :o)
Aha! I updated one node.js version later (0.11.14), and that solved the problem. No more vulnerabilities for me! :o)
One of the projects requires support much higher version php used in main web project
Env:
- Debian 7.6 (3.2 x86_64)
- Nginx 1.6.0
- php-fpm (PHP 5.3.28)
My task: Install alternate version of php-fpm (php) and use it in another web project.
I don't want to install Apache with the desired version of php (mod-php) and use Nginx as frontend for Apache
Please describe other ways of resolving my situation, if they exist
I was researching this topic in the past. Never went through with it but you might find this useful.