/// article

Troubleshooting Zimbra Mailboxd When the Service Will Not Start

When Zimbra mailboxd refuses to start, the problem is often deeper than the service itself. This guide shows how to trace the failure through logs, MySQL, Java, ports, permissions, storage, certificates, and other dependencies. When Zimbra reports: mailbox Stopped zmmailboxdctl is not running the natural reaction is often to run: zmmailboxdctl restart aga...

By lordfrancs3

When Zimbra mailboxd refuses to start, the problem is often deeper than the service itself. This guide shows how to trace the failure through logs, MySQL, Java, ports, permissions, storage, certificates, and other dependencies. When Zimbra reports: mailbox Stopped zmmailboxdctl is not running the natural reaction is often to run: zmmailboxdctl restart again and again. That rarely helps if mailboxd is failing because one of its dependencies is unavailable or its configuration is damaged. During our Zimbra recovery, zmmailboxdctl would not operate normally while the server was also experiencing database problems. MySQL could not start normally, InnoDB required emergency recovery, database dumps were failing, and later database authentication problems appeared after database initialization work. In that situation, mailboxd itself was not necessarily the root problem. The correct approach was to investigate the dependencies beneath mailboxd. Important: A stopped mailbox service can have many causes. Do not assume that a MySQL problem, certificate problem, port conflict, or permission problem is responsible until the logs and system state support that conclusion. What Is Zimbra Mailboxd? Mailboxd is one of the core Zimbra services. It runs the mailbox application and handles functions associated with user mailboxes, the web application, administration services, LMTP delivery, and several backend protocols. Zimbra’s service controller reports its status as: su - zimbra zmcontrol status A failure may appear similar to: Host mail.example.com ldap Running mailbox Stopped zmmailboxdctl is not running mta Running Zimbra documents zmcontrol status as the command for displaying the state of Zimbra components. The official reference is: https://wiki.zimbra.com/wiki/Zmcontrol What Happened in Our Recovery Our mailboxd problem occurred during a much larger recovery. The Zimbra database was damaged and would not start normally. We attempted InnoDB forced recovery, and eventually: innodb_force_recovery = 6 allowed MySQL to start sufficiently for database inspection. We could then run: SHOW DATABASES; and see the Zimbra databases. However, logical dump attempts still encountered problems. Later, database initialization work using: /opt/zimbra/libexec/zmmyinit changed the database environment, and we subsequently encountered database authentication problems involving the Zimbra database user. At the same time: zmmailboxdctl is not running remained part of the recovery problem. This gave us an important troubleshooting principle: Before troubleshooting mailboxd as an isolated Java service, verify that the database and other services it depends on are healthy. Do Not Start by Repeatedly Restarting Mailboxd A restart is useful once. It can reproduce the problem and create fresh log entries. For example: su - zimbra zmmailboxdctl start Then check: zmcontrol status If mailboxd immediately stops again, do not keep running: zmmailboxdctl restart without investigating the reason. Repeated restarts can: overwrite useful startup context generate large amounts of repetitive log data complicate timestamps place additional load on a damaged database hide the original sequence of events The next step should be log inspection. Step 1: Establish the Overall Zimbra State Switch to the Zimbra user: su - zimbra Run: zmcontrol status Do not look only at: mailbox Check the complete result. For example: ldap Running mailbox Stopped mta Running proxy Running is different from: ldap Stopped mailbox Stopped mta Stopped The first situation suggests a more localized mailbox problem. The second suggests a broader dependency or configuration failure. You can also record the Zimbra version: zmcontrol -v This is useful because command paths, Java versions, database components, ports, and configuration behavior differ across Zimbra releases. Step 2: Check zmmailboxd.out First When mailboxd will not start, one of the most useful logs is: /opt/zimbra/log/zmmailboxd.out Zimbra specifically documents this file as containing mailboxd startup output and recommends reviewing it when mailboxd fails to start. Start with: tail -100 /opt/zimbra/log/zmmailboxd.out For a larger view: tail -300 /opt/zimbra/log/zmmailboxd.out Search for common failure terms: grep -iE \ 'error|exception|failed|fatal|denied|bind|address already|keystore|certificate|mysql|ldap|java|memory' \ /opt/zimbra/log/zmmailboxd.out | tail -100 This file can expose errors that never reach mailbox.log because mailboxd fails before the application finishes initializing. Official reference: https://wiki.zimbra.com/wiki/Troubleshooting_Course_Content_Rough_Drafts-Zimbra_Logs Step 3: Inspect mailbox.log The primary mailbox application log is: /opt/zimbra/log/mailbox.log Zimbra documents this as the main log for mailboxd activity. Review recent events: tail -200 /opt/zimbra/log/mailbox.log Search for serious errors: grep -iE \ 'fatal|error|exception|mysql|sql|ldap|certificate|keystore|permission|denied|connection refused' \ /opt/zimbra/log/mailbox.log | tail -100 The most useful approach is to reproduce the startup failure while watching the logs. In one terminal: tail -f /opt/zimbra/log/zmmailboxd.out In another: tail -f /opt/zimbra/log/mailbox.log Then: zmmailboxdctl start Watch which error appears first. The first meaningful error is often more useful than the final cascade of exceptions. Step 4: Check MySQL Before Blaming Mailboxd Zimbra’s mailbox troubleshooting documentation identifies database connectivity as a common cause of mailbox startup problems. This was particularly relevant in our recovery. Check whether the database is running: mysql.server status Depending on your Zimbra release, the service-control details may vary, but the principle is the same. Try connecting using the Zimbra environment: mysql zimbra or: mysql mysql If you receive something like: ERROR 1045 (28000): Access denied for user ... you have discovered a database authentication problem. If you receive: Can't connect to local MySQL server then mailboxd may simply have no functioning database to connect to. Zimbra’s mailbox troubleshooting guide specifically recommends testing database access when mailbox fails to start: https://wiki.zimbra.com/wiki/Mailbox_Troubleshooting Step 5: Inspect mysql_error.log Database startup problems should be investigated in: /opt/zimbra/log/mysql_error.log Zimbra documents this as the log containing MySQL startup and database health information, including InnoDB corruption errors. Check: tail -200 /opt/zimbra/log/mysql_error.log Search for: grep -iE \ 'error|innodb|corrupt|crash|tablespace|page|denied|password|failed' \ /opt/zimbra/log/mysql_error.log | tail -100 During our recovery, the database problem was serious enough that normal MySQL startup failed. Mailboxd troubleshooting therefore could not reasonably proceed as though MySQL were healthy. A Database Dependency Failure Can Look Like a Mailbox Failure Consider this simplified dependency chain: mailboxd | +---- MySQL / MariaDB | +---- LDAP | +---- Java runtime | +---- filesystem | +---- certificates / keystore | +---- network ports If MySQL is unavailable: MySQL fails | v Mailboxd cannot initialize correctly | v zmmailboxdctl is not running Restarting mailboxd does not repair the database. This was one of the central lessons from our incident. Step 6: Verify the Database Password Configuration A mismatch between the database password stored by Zimbra and the actual database account can prevent mailboxd from accessing its database. Zimbra’s troubleshooting documentation specifically identifies an out-of-sync MySQL password as a possible cause of mailbox startup failure. You can inspect relevant local configuration as the Zimbra user. Be careful because the following command can expose secrets: zmlocalconfig -s | grep mysql Do not paste the output into tickets, forums, blog comments, or public chats. A safer initial inspection is: zmlocalconfig | grep mysql If you need to investigate the actual stored password, handle it as sensitive data. Our recovery experienced exactly this type of complication after database initialization work changed database credentials. Step 7: Check LDAP Mailboxd also depends heavily on LDAP. Check overall status: zmcontrol status Or check LDAP directly using the commands available for your installation. If LDAP is stopped, investigate it before expecting mailboxd to initialize normally. Look for LDAP-related errors in: /opt/zimbra/log/mailbox.log /var/log/zimbra.log A typical dependency failure might appear as: LDAP error connection refused unable to bind TLS negotiation failed Certificate problems can also prevent LDAP TLS connections, so an apparent LDAP problem may eventually lead back to certificates. Step 8: Check Disk Space Mailboxd, MySQL, Jetty, Java, logs, indexes, and temporary operations all require disk space. Check: df -h Pay particular attention to: / /opt /opt/zimbra /tmp /var/tmp if they are on separate filesystems. A filesystem at: 100% can prevent: Java startup database writes log creation temporary-file creation mailbox indexing database transactions PID-file creation Also check inode exhaustion: df -i A filesystem can have gigabytes of free storage but still be unable to create files if all inodes are consumed. Step 9: Check the Filesystem Mount State Disk capacity is not the only storage concern. Check: findmnt / and: findmnt /opt/zimbra Look for: rw versus: ro A filesystem that has been remounted read-only can make services fail even when: df -h looks normal. During our broader Zimbra recovery, the underlying EXT4 filesystem had previously experienced read-only behavior, making filesystem health an important part of the investigation. Search the kernel log: dmesg -T | grep -Ei \ 'ext4|I/O error|filesystem|read-only|remount|buffer error' On systemd-based Ubuntu systems, you can also inspect: journalctl -k Look for storage or filesystem failures before assuming the problem belongs to Zimbra. Step 10: Check /tmp Applications frequently depend on temporary directories. Check: ls -ld /tmp Typical Linux /tmp permissions normally include the sticky bit: drwxrwxrwt Also check filesystem attributes: lsattr -d /tmp During our troubleshooting, we found: ----ia-------e-- /tmp The important attributes were: i = immutable a = append-only Those attributes can restrict normal filesystem operations even when ordinary Unix permissions appear writable. Therefore, if mailboxd or Java reports: Operation not permitted Cannot create temporary file Permission denied do not repeatedly change permissions without understanding the filesystem state. Check: ls -ld /tmp lsattr -d /tmp getfacl -p /tmp findmnt -T /tmp df -h /tmp df -i /tmp Step 11: Check Permissions Under /opt/zimbra Incorrect ownership or permissions can prevent Zimbra components from reading configuration, accessing certificates, creating files, or starting services. Inspect relevant paths: ls -ld /opt/zimbra ls -ld /opt/zimbra/log ls -ld /opt/zimbra/store ls -ld /opt/zimbra/index Do not immediately run: chown -R zimbra:zimbra /opt/zimbra That is dangerous because not every Zimbra file should necessarily have identical ownership. Zimbra provides: /opt/zimbra/libexec/zmfixperms for repairing expected permissions under /opt/zimbra. Run it as root when permissions are a confirmed or reasonable suspected issue: /opt/zimbra/libexec/zmfixperms Zimbra also provides: /opt/zimbra/libexec/zmfixperms -extended for a broader check that includes more data paths, but Zimbra warns this can take substantially longer on large environments. Official reference: https://wiki.zimbra.com/wiki/Fix_the_Zimbra_Collaboration_Permissions Step 12: Check Java and JVM Configuration Mailboxd runs as a Java application. If the JVM cannot start, mailboxd cannot start. A first check is: ps -ef | grep '[j]ava' If mailboxd is stopped, you may find no corresponding Java process. Inspect configured mailboxd JVM options: zmlocalconfig mailboxd_java_options Depending on the Zimbra release, heap configuration can also be inspected using variables such as: zmlocalconfig mailboxd_java_heap_size Zimbra documents mailboxd_java_options as the local configuration controlling JVM options for mailboxd. It also warns that an invalid Java option can prevent the mailbox Java process from starting. Do not blindly copy JVM options from another server or an old forum post. Java flags that were valid for an older JVM may have been removed in a newer JVM. If someone recently modified: mailboxd_java_options and mailboxd stopped afterward, compare the current value against the expected value for the installed Zimbra release. Look for Java-Specific Errors Search: grep -iE \ 'java|jvm|heap|outofmemory|unsupported|unrecognized|could not create' \ /opt/zimbra/log/zmmailboxd.out | tail -100 Common categories include: OutOfMemoryError Could not create the Java Virtual Machine Unrecognized VM option Unable to allocate memory Java keystore errors If the problem is memory-related, also inspect the operating system: free -h and: ps aux --sort=-%mem | head -20 Do not assume that adding more Java heap memory is always the solution. An oversized heap can itself prevent Java from starting if sufficient address space or physical memory is unavailable. Step 13: Check Ports Mailboxd needs to bind to several ports depending on your Zimbra topology and configuration. Zimbra documents common mailbox-related internal ports such as: 7025 LMTP 7071 Admin HTTPS 7072 Nginx lookup 7073 SASL lookup 7306 database 8080 backend HTTP 8443 backend HTTPS with additional IMAP and POP backend ports depending on configuration. Official reference: https://wiki.zimbra.com/wiki/Ports Check listening ports: ss -lntp For specific ports: ss -lntp | grep -E \ ':(7025|7071|7072|7073|7306|8080|8443)\b' Do not assume that every port listed above must be active on every server. Zimbra roles and proxy configuration affect which ports are used. Look for Port Conflicts A mailbox service can fail if another process already owns the port it needs. For example: ss -lntp | grep ':8080' might reveal another application listening there. Check the process: ps -fp <PID> Zimbra’s own mailbox troubleshooting documentation notes that web servers or other services occupying required ports can prevent mailbox startup. Do not immediately kill the competing process. First determine: what it is why it is running whether Zimbra is actually configured for that port whether the conflict appeared after an upgrade or configuration change Step 14: Verify Zimbra Port Configuration Rather than assuming defaults, inspect the configured values. For example: zmprov gs "$(zmhostname)" | grep -Ei \ 'zimbraMailPort|zimbraMailSSLPort|zimbraAdminPort|zimbraLmtpBindPort' The exact attributes vary by release. The purpose is to answer: Which ports is this particular mailbox server supposed to bind? Compare the expected configuration with: ss -lntp That is much better than assuming every Zimbra installation uses the same network layout. Step 15: Check Configuration Changes Mailboxd may fail after an incorrect Zimbra configuration value is written. Zimbra documents cases where incorrect server settings, including bind addresses, can prevent mailbox startup. Useful commands include: zmlocalconfig and: zmprov gs "$(zmhostname)" Do not dump full configuration into a public ticket because it may expose: internal hostnames passwords tokens LDAP information database details Instead, search for the configuration area associated with the error. For example: zmprov gs "$(zmhostname)" | grep -i bind or: zmlocalconfig | grep -i mailbox If the failure started immediately after a configuration change, compare with a known-good backup or configuration-management history. Step 16: Check the Certificate and Java Keystore Certificates are not the first thing to blame every time mailboxd fails. However, mailboxd uses Java keystores and TLS, so certificate or keystore failures can prevent startup. Zimbra documents a mailbox startup failure containing: java.io.IOException: Keystore was tampered with, or password was incorrect as a certificate or keystore-related problem. Search: grep -iE \ 'keystore|certificate|ssl|tls|pkix|trust|expired' \ /opt/zimbra/log/zmmailboxd.out | tail -100 and: grep -iE \ 'keystore|certificate|ssl|tls|pkix|trust|expired' \ /opt/zimbra/log/mailbox.log | tail -100 Inspect the Deployed Certificate Zimbra’s zmcertmgr can display the certificate used by mailboxd. As the Zimbra user: /opt/zimbra/bin/zmcertmgr viewdeployedcrt mailboxd You can also check expiration: /opt/zimbra/bin/zmcertmgr checkcrtexpiration mailboxd Zimbra documents both commands in its certificate-management tools. Official reference: https://wiki.zimbra.com/wiki/Zmcertinstall Verify a Certificate Before Deploying Anything If certificate files are being investigated, Zimbra supports: /opt/zimbra/bin/zmcertmgr verifycrt \ comm \ <PRIVATE_KEY> \ <CERTIFICATE> \ <CA_CHAIN> This checks both: certificate versus private key certificate chain validity Do not redeploy certificates simply because mailboxd is down. First confirm that the logs actually indicate a certificate problem. A bad database will not be repaired by regenerating certificates. Step 17: Check LDAP TLS Failures Certificates can also break dependencies rather than mailboxd directly. For example, an expired root or intermediate CA can prevent LDAP TLS communication. Zimbra documents errors such as: Unable to start TLS SSL connect attempt failed when certificate chains expire. If mailboxd cannot communicate with LDAP because TLS verification fails, you may see the mailbox service fail even though its Java process initially starts. This is why dependency troubleshooting matters. Step 18: Check Hostname and Name Resolution Zimbra relies heavily on its hostname and LDAP configuration. Check: hostname hostname -f and: zmhostname The values should make sense for the installed environment. Check resolution: getent hosts "$(zmhostname)" If the server’s hostname suddenly resolves to the wrong address, mailboxd may fail to reach local or remote dependencies correctly. Do not randomly edit: /etc/hosts /etc/hostname on a production Zimbra server without understanding the existing DNS and Zimbra topology. Step 19: Check Memory and Resource Exhaustion Mailboxd can also fail because the server lacks memory. Check: free -h Then: vmstat 1 5 and: ps aux --sort=-%mem | head -20 Look for: very little available RAM heavy swapping another process consuming most memory an oversized mailboxd Java heap OutOfMemoryError kernel OOM events Check kernel events: dmesg -T | grep -iE \ 'out of memory|oom|killed process' If Java is being killed by the kernel’s OOM killer, zmmailboxdctl restart may appear to work briefly and then mailboxd will disappear again. Step 20: Read the Logs Together Do not investigate each log in isolation. A useful terminal layout is: tail -f /opt/zimbra/log/zmmailboxd.out tail -f /opt/zimbra/log/mailbox.log tail -f /opt/zimbra/log/mysql_error.log and, when useful: tail -f /var/log/zimbra.log Then reproduce one startup attempt. For example: zmmailboxdctl start Now correlate events by time. You may see: 09:10:04 mailboxd Java starts 09:10:06 database connection fails 09:10:07 application initialization aborts 09:10:08 mailboxd exits That sequence tells you much more than: zmmailboxdctl is not running Actual Findings From Our Recovery The following were confirmed during our recovery: mailboxd was not running correctly MySQL would not start normally InnoDB corruption required forced-recovery investigation lower recovery levels did not provide a usable database innodb_force_recovery=6 eventually allowed database inspection Zimbra databases were visible with SHOW DATABASES some database dump attempts failed database initialization was later performed database authentication problems appeared during the recovery the server had also experienced filesystem problems Given those conditions, it would have been misleading to say: mailboxd itself was definitely broken. A better conclusion was: mailboxd could not be trusted to start normally while several required dependencies were unhealthy. Additional Checks Recommended in This Guide The following checks are general troubleshooting techniques. They were not necessarily all responsible for our specific mailboxd outage: port conflicts Java heap problems invalid JVM options certificate expiration Java keystore corruption LDAP TLS problems incorrect bind addresses wrong mailbox port configuration disk-full conditions inode exhaustion incorrect Zimbra permissions hostname resolution problems OOM killer activity These should be investigated when the logs point in their direction. Do not assume every mailboxd failure has the same root cause. A Recommended Troubleshooting Sequence When mailboxd will not start, I recommend working through the problem in this order. 1. Capture the current state su - zimbra zmcontrol -v zmcontrol status Record the output. 2. Attempt one controlled start zmmailboxdctl start Do not keep restarting it repeatedly. 3. Read mailboxd startup output tail -200 /opt/zimbra/log/zmmailboxd.out 4. Read mailbox application errors tail -200 /opt/zimbra/log/mailbox.log 5. Verify MySQL mysql.server status mysql zimbra Then inspect: tail -200 /opt/zimbra/log/mysql_error.log 6. Verify LDAP and other Zimbra services zmcontrol status 7. Check storage df -h df -i findmnt 8. Check filesystem errors dmesg -T | grep -Ei \ 'I/O error|ext4|filesystem|read-only|remount' 9. Check /tmp ls -ld /tmp lsattr -d /tmp getfacl -p /tmp 10. Check Java configuration zmlocalconfig mailboxd_java_options and: free -h 11. Check ports ss -lntp 12. Check permissions if indicated As root: /opt/zimbra/libexec/zmfixperms 13. Check certificates if the logs indicate TLS or keystore errors /opt/zimbra/bin/zmcertmgr \ viewdeployedcrt mailboxd 14. Correct the root cause Only after identifying it. 15. Start mailboxd again zmmailboxdctl start Then: zmcontrol status A Simple Decision Tree A practical way to think about mailboxd failure is: mailboxd stopped | v Check zmmailboxd.out | +---- Database error? | | | v | Fix MySQL/database | +---- LDAP error? | | | v | Fix LDAP/TLS/DNS | +---- Bind/port error? | | | v | Check ss and Zimbra port config | +---- Java/JVM error? | | | v | Check Java options and memory | +---- Permission error? | | | v | Inspect ownership, then zmfixperms | +---- Keystore/TLS error? | | | v | Inspect certificates and keystore | +---- Filesystem error? | v Check disk, inode, mounts, /tmp and kernel logs This is much more effective than: restart restart restart restart When zmfixperms Is Appropriate Run: /opt/zimbra/libexec/zmfixperms when evidence points to incorrect Zimbra ownership or permissions. Do not use it as the first response to every mailbox failure. Likewise, use: /opt/zimbra/libexec/zmfixperms -extended only when broader permission checking is justified. The extended operation can take a long time on systems with large stores and indexes. Zimbra explicitly warns about this. When Certificate Repair Is Appropriate Certificate repair is justified when the logs contain errors such as: Keystore was tampered with certificate expired unable to validate certificate chain PKIX path building failed TLS negotiation failed Do not regenerate the CA or deploy a new certificate just because: mailboxd stopped Changing certificates unnecessarily during an unrelated recovery can create a second problem. When MySQL Should Take Priority If you see: Can't connect to local MySQL server Access denied for user InnoDB corruption tablespace corruption database crash recovery failure stop troubleshooting mailboxd as though it were independent. Fix or recover the database first. Zimbra’s MySQL crash-recovery documentation follows the same logic. It instructs administrators to restore and test the database before starting the remaining Zimbra services, and then review mysql_error.log and mailbox.log for database errors. Official reference: https://wiki.zimbra.com/wiki/Mysql_Crash_Recovery Do Not Ignore Dependency Order In a damaged environment, start with foundational services. A simplified dependency order is: Operating system | v Filesystem and storage | v Networking and DNS | v LDAP | v Database | v Java / mailboxd | v Webmail / admin / mailbox protocols The exact Zimbra architecture varies, especially in multi-server deployments, but the principle remains useful. A failure lower in the stack can appear as a mailboxd problem higher in the stack. Be Extra Careful After Database Recovery If mailboxd previously failed because the database was corrupted, do not simply start all services immediately after MySQL accepts connections. First test: mysql -e "SHOW DATABASES;" Then inspect important databases: mysql zimbra -e "SHOW TABLES;" Review: tail -100 /opt/zimbra/log/mysql_error.log Only after the database behaves normally should you attempt: zmmailboxdctl start This is particularly important after using: innodb_force_recovery because forced recovery is intended for data extraction, not normal production operation. Collect Diagnostics Before Making Major Changes If the cause remains unclear, Zimbra provides diagnostic collection tools. One documented command is: /opt/zimbra/libexec/zmdiaglog Zimbra documents this as a way to collect relevant troubleshooting information. Be aware that diagnostic archives can contain: internal hostnames configuration information logs email addresses infrastructure details Treat them as sensitive. Do not upload them to public forums without reviewing and sanitizing them. Lessons Learned From Our Recovery The first lesson was that mailboxd being stopped does not mean mailboxd is the root cause. Our server had a damaged database. Restarting the Java application could not repair InnoDB. The second lesson was that startup logs matter more than the service-status line. zmmailboxd.out can explain why mailboxd failed before mailbox.log even had a chance to record normal activity. The third lesson was that dependencies should be checked in order. A broken filesystem can break MySQL. Broken MySQL can break mailboxd. Mailboxd then becomes the visible symptom. The fourth lesson was that recovery work can introduce new problems. Database initialization, credential changes, permission changes, and certificate changes must be documented carefully. The fifth lesson was that ports and configuration should be verified rather than assumed. Zimbra configurations vary between single-server, proxy, and multi-server deployments. The sixth lesson was that permissions should be repaired using Zimbra’s own tools where possible, rather than broad recursive chmod or chown commands. Finally, restart is a test, not a troubleshooting strategy. A useful restart reproduces the failure and gives you logs. After that, investigate the reason. Quick Troubleshooting Checklist When you see: zmmailboxdctl is not running check: [ ] zmcontrol status [ ] Zimbra version [ ] zmmailboxd.out [ ] mailbox.log [ ] mysql_error.log [ ] MySQL connectivity [ ] LDAP status [ ] disk space [ ] inode availability [ ] filesystem read/write status [ ] kernel filesystem errors [ ] /tmp permissions and attributes [ ] mailboxd Java options [ ] available memory [ ] OOM events [ ] listening ports [ ] Zimbra port configuration [ ] /opt/zimbra ownership and permissions [ ] certificate and keystore errors [ ] hostname and DNS resolution Then correct the specific failure that the evidence identifies. References Zimbra Mailbox Troubleshooting Zimbra’s mailbox troubleshooting article covers database connectivity, service startup problems, and examples of zmmailboxdctl not running. https://wiki.zimbra.com/wiki/Mailbox_Troubleshooting Zimbra Log File Reference Documents /opt/zimbra/log/mailbox.log, /opt/zimbra/log/zmmailboxd.out, /opt/zimbra/log/mysql_error.log, and other major troubleshooting logs. https://wiki.zimbra.com/wiki/Troubleshooting_Course_Content_Rough_Drafts-Zimbra_Logs Zimbra Performance Troubleshooting Provides additional information about mysql_error.log, zmmailboxd.out, JVM behavior, and mailboxd troubleshooting. https://wiki.zimbra.com/wiki/Troubleshooting_Performance_Issues Zimbra zmcontrol Documents zmcontrol status, start, stop, and other Zimbra service-control commands. https://wiki.zimbra.com/wiki/Zmcontrol Zimbra Port Reference Lists Zimbra service ports, including mailbox, database, LMTP, admin, backend HTTP, and HTTPS ports. https://wiki.zimbra.com/wiki/Ports Fix Zimbra Permissions Documents the supported zmfixperms and zmfixperms -extended procedures. https://wiki.zimbra.com/wiki/Fix_the_Zimbra_Collaboration_Permissions Zimbra Certificate Tools Documents zmcertmgr, including certificate verification, deployed-certificate inspection, and expiration checking. https://wiki.zimbra.com/wiki/Zmcertinstall Zimbra MySQL Crash Recovery Documents database recovery, restoring the Zimbra database before mailbox groups, and checking mysql_error.log and mailbox.log before returning services to operation. https://wiki.zimbra.com/wiki/Mysql_Crash_Recovery SEO and Social Metadata SEO Title: Troubleshooting Zimbra Mailboxd When the Service Will Not Start SEO Description: Learn how to troubleshoot a stopped Zimbra mailboxd service by checking logs, MySQL, Java, ports, permissions, storage, certificates, LDAP, and dependencies. Social Title: Zimbra Says “zmmailboxdctl Is Not Running”? Check These First Social Description: Restarting mailboxd repeatedly will not fix a broken database, port conflict, Java error, filesystem problem, or certificate issue. Follow this practical Zimbra troubleshooting sequence instead. Focus Keyword: zmmailboxdctl is not running Tags: Zimbra, Zimbra Mailboxd, zmmailboxdctl, Ubuntu Server, Linux Administration, Zimbra Troubleshooting, MySQL, Java, Mail Server, Zimbra Recovery, Server Troubleshooting, System Administration