| Change | User | Description | Committed |
|---|---|---|---|
| 31345 | Mark Zinthefer | Adding trigger script to enforce the type of a changelist to "restricted" | 4 days ago |
| 31329 | C. Thomas Tyler | Added summary of errors and warnings at the end. This summary is displayed by default. A... '-skip_summary' option has been added to suppress the summary. Added checks for existence of statepullL and statefailover files in P4ROOT on p4d servers. A warning is reported if these files exist, advising the user to remove these failover remnant files. Addressed some issues reported by shellcheck. #review-31330 @robert_cowham « |
8 days ago |
| 31317 | C. Thomas Tyler | Tweaked set of tables to be used for edge servers to work for recovery with modern SDP on... a pre-2019.1 data set. The CheckpointTables variable appends db.storage only if the p4d version is new enough to use db.storage. « |
21 days ago |
| 31311 | C. Thomas Tyler | Updated run_if_*.sh scripts: * Fixed bug in run_if_broker.sh. * Normalized command line v...erification. * Removed 'set -e' from run_if_*.sh scripts to improve reliability. * Made ShellCheck compliant. #review-31312 « |
about a month ago |
| 31304 | C. Thomas Tyler | Corrected '*/60 * * * *' entries in crontab to be "0 * * * *'. This has the same ... meaning (run every hour), but uses valid syntax that works across Linux distros. The */60 syntax, while technically invalid, is interpreted as intended on RHEL/Rocky and Ubuntu distros, but is rejected as invalid on SuSE Linux (where it is righteously rejected). Replacing '*/60' with '0' has the same symantic mealing but uses a syntax valid across platforms. « |
about a month ago |
| 31243 | C. Thomas Tyler | Fixed typo in error message. | 2 months ago |
| 31210 | C. Thomas Tyler | Fixed issue with case-sensitivty of server spec name in configurable values. For example,... if the server spec name is p4d_edge_SITEX and is case-insensitive, the configurables displayed with p4d -cshow can look like P4D_EDGE_sitex and still be valid. An example of an issue this fixes that an edge server being migrated from Windows to Linux could have the run_if_edge.sh script fail to determine that the server is an edge server; a daily_checkpoint.sh would run fine interactively but would fail to execute from the crontab, due to the run_if_edge.sh not detecting it was an edge. The situation that could cause this would be that a ServerID could be p4d_edge_SYD but the lbr.replication configurable could be set as P4D_EDGE_SYD#lbr.replication=readonly #review-31211 « |
2 months ago |
| 31209 | C. Thomas Tyler | Fixed cosmetic typo. | 2 months ago |
| 31185 | C. Thomas Tyler | Added more detailed information on The Site Directory (/p4/common/site). Adapted Makefi...le to generate HTML from Markdown for this special case where the /p4/common/site/ReadMe.md file deployed with the SDP in a runtime environment in the site folder generates an HTML elsewhere, in the SDP doc folder. Fixes SDP-722. « |
2 months ago |
| 31182 | C. Thomas Tyler | Adapted opt_perforce_sdp_backup service to work with SELinux in enforcing mode. Added... logic from sdp_upgrade.sh that installs SDP the backup service into install_sdp.sh. With this adjustment, the service now specifies the "on disk" location of the file; no symlinks are allowed in the absolute path to the script (to keep SELinux happy). For clarity, 'User=root' was added to the systemd service file for the backup utility. « |
2 months ago |
| 31178 | C. Thomas Tyler | Tweaked error message in sdp_upgrade.sh; non-functional change. | 2 months ago |
| 31170 | C. Thomas Tyler | In verify_sdp.sh, added '-extra' to specify tests not normally run by default, but which... can be called on request. Two additional tests added are: * commit_defined * server_type_known Fixes: * SDP-1187 * SDP-1188 #review-31171 « |
2 months ago |
| 31167 | C. Thomas Tyler | Fixed p4login bug for case-insensitive servers checking for SERVERID using case case-sens...itive check. Fixes SDP-1189 (Bug): p4login is wrongly sensitive to case of ServerID on case-insensitive server. « |
2 months ago |
| 31158 | C. Thomas Tyler | This is a change to unreleased new behaviour to handle the clients.readonly.dir data as p...art of the database swap handling. Added a disk space safety check, so the refresh_P4ROOT_from_offline_db will safely abort early if there's not enough disk space to safely complete the rsync of the readonly.clients.dir folder to the offline_db prior to the database swap orchestrated by refresh_P4ROOT_from_offline_db.sh. Renamed the copy function for improved clarity, and added a new rsync_with_preflight function to improve modularity. #review-31159 « |
2 months ago |
| 31152 | C. Thomas Tyler | Added feature to preserve client.readonly.dir during a database swap. Refined tmp file... cleanup logic to search only for tmp files owned by the defined OSUSER. #review-31153 @robert_cowham « |
2 months ago |
| 31136 | C. Thomas Tyler | Added content to cover scenario where P4JOURNAL is set in db.config. This addresses SDP...-737 (Doc): In SDP Legacy Upgrade Guide, advise p4d -cunset P4JOURNAL if needed. Also updated URL for list of configurables due to change in Perforce web site layout, with docs moved from somewhere under www.perforce.com to somewhere under help.perforce.com. For example: Old URL for configurables: https://www.perforce.com/manuals/cmdref/Content/CmdRef/configurables.alphabetical.html New URL for configurables: https://help.perforce.com/helix-core/server-apps/cmdref/current/Content/CmdRef/configurables.alphabetical.html « |
2 months ago |
| 31108 | C. Thomas Tyler | Changed the SDPInfo trigger to display the additional output only if the client program c...alling the trigger is 'p4'. This fixes an issue where installing the trigger causes a Swarm crash. « |
2 months ago |
| 31101 | C. Thomas Tyler | Fixed bug where a case-insensitive server is sensitive to the case of a server spec in th...e server.id file. If the instance is case-sensitive, then a grep for the ServerID in dumped db.server output should also be case-insensitive. An impact of the bug is that the run_if_*.sh scripts, e.g. run_if_edge.sh, will not work if the case of the ServerID in the $P4ROOT/server.id file doesn't match what shows in 'p4 servers'. The case must match if the data set is case- sensitive, but is not required to match if it is insensitive (but due to the bug, is must match). Without this code change in SDP, an easy workaround for this issue is simply to adjust the case of the ServerID value in the $P4ROOT/server.id file match what appears in the output of 'p4 servers'. This bug may be more likely to appear in a Windows to Linux migration. Fixes SDP-1184. #review-31102 @leif.johnson « |
3 months ago |
| 31082 | C. Thomas Tyler | Fixed bug with sdp_upgrade.sh munging values during upgrade for P4USER, etc. Fixed SDP-...1183. « |
3 months ago |
| 31051 | C. Thomas Tyler | Tweaked to create backup dir if needed. | 3 months ago |
| 31050 | C. Thomas Tyler | Added script to back SDP OS Package Structure. Modified sdp_upgrade.sh to install a sys...temd oneshot service to call this script if operating on a system using the SDP OS Package Structure (i.e. a systemd where /opt/perforce/helix-sdp exists). « |
3 months ago |
| 31048 | C. Thomas Tyler | Adjustment to create `/opt/perforce/helix-sdp/backup` directory. | 3 months ago |
| 31045 | C. Thomas Tyler | Terminology change; now referring to two structures as "SDP Classic Structure" and "SDP O...S Package Structure". « |
3 months ago |
| 31041 | C. Thomas Tyler | daily_checkpoint.sh on edge skips checkpoint dump if journal rotation needed. Prior beh...avior was to fail with an error. However, this situation comes up routinely after seeding or reseeding an edge server, and is not reallly an error condition. With the new logic, the iteration of the checkpoint dump is simply skipped. The next call made after a journal rotation on the commit will succeed. If the checkpoint dump is skipped, calls to recreate_offline_db_files(), remove_old_checkpoints_and_journals(), remove_old_logs() all just display a message indicating they do nothing because the checkpoint dump was skipped. #review-31042 « |
3 months ago |
| 31039 | C. Thomas Tyler | Shellcheck cleanup. | 3 months ago |
| 31033 | C. Thomas Tyler | sdp_upgrade.sh: Refined logic extracting old values during upgrades. | 3 months ago |
| 31028 | C. Thomas Tyler | In verify_sdp.sh, changed call to check_dir_ownership.sh to use faster test, matching wha...t is required by p4d_base to start the service. « |
3 months ago |
| 31023 | C. Thomas Tyler | The live_checkpoint.sh script now works on replicas and edges. TO DO: * Review docs fo...r any needed updates. * Add test suite coverage. Fixes SDP-1079. #review-31024 @d_benedict « |
3 months ago |
| 31022 | C. Thomas Tyler | Removed calls to remove_old_checkpoints_and_journals and remove_old_logs from request_rep...lica_checkpoint.sh. The removal of old checkpoints/logs is only to be done upon successful completion of a new checkpoint operation, and this script only makes the request to start a checkpoint; it does not verify successful completion. « |
3 months ago |
| 31021 | C. Thomas Tyler | Refined to avoid possible delays if scanning P4ROOT with 'find' if P4ROOT contains large... subdirs (server.locks, etc.). The 'find' is restricted with a maxdepth of 1. « |
3 months ago |
| 31020 | C. Thomas Tyler | Follow on change for @31019. | 3 months ago |
| 31018 | C. Thomas Tyler | Renamed unreleased new script, check_dir_perms.sh -> check_dir_ownership.sh. Fixed exi...t code in this new script. Modified p4d_base to call this script as a preflight check when starting p4d (regular start only, not force_start). It will refuse to start p4d if there are any files in P4ROOT owned by a user other than the defined OSUSER. This should be helpful in finding things like a root-owned state.xu file. Fixes SDP-1119. #review-31019 « |
3 months ago |
| 31015 | C. Thomas Tyler | Adjusted 'find' command to eliminate redundant second '-mtime' option. | 3 months ago |
| 31010 | C. Thomas Tyler | Fixed issue where values containing equal signs (e.g. PROXY_V_FLAGS using '-v name=value...' settings) resulted in a bad substitution. « |
3 months ago |
| 31000 | C. Thomas Tyler | Added SDP_ROOT setting for future usage. | 3 months ago |
| 30984 | C. Thomas Tyler | Enhanced sdp_upgrade.sh to work with both the new Package Structure /opt/perforce/helix-s...dp and the legacy SDP structure when doing upgrades of /p4/common and /p4/sdp. If the Package Structure is already used, this will continue using it. If the legacy strucutre is in use, it will continue using that. As the helix-sdp package is not yet available, the Package Structure will exist only if the new install_sdp.sh script was used for the install (or if the customer is one of a few early adopters of the new structure). This version, to be released with SDP 2024.2, will NOT migrate from the legacy structure to the Package structure. Doing so requires running as root; that change is targeted for release in SDP 2025.1. Also removed the dependency on perl from sdp_upgrade.sh. Addressed suggestions from newer version of ShellCheck. #review-30985 « |
3 months ago |
| 30979 | C. Thomas Tyler | Eliminated buildup of temp dirs, e.g. /tmp/tmp.XXXXXXXXXX. Added remove_jd_tables() fu...nction and calls to it to prevent buildup of new cruft. Modified remove_old_logs() to cleanup cruft created previously. #review-30980 @robert_cowham « |
3 months ago |
| 30950 | John Clucas | Fix python3 issues with map | 3 months ago |
| 30937 | Robert Cowham | Update p4review2.py to work with Python3 Add basic test harness. Delete p4review.py whic...h is Python2 and update docs. « |
4 months ago |
| 30933 | C. Thomas Tyler | Added check for proper ownership of all files in P4ROOT. Added separate utility for che...cking ownership (and optionally group) of all files in a given path. Fixes SDP-1119. #review-30934 « |
4 months ago |
| 30924 | C. Thomas Tyler | Fixed an obscure "one day a month" bug that caused sdp_upgrade.sh to fail to detect the S...DP version from the version file if released on the 20th of any given month. « |
4 months ago |
| 30911 | C. Thomas Tyler | Adjusted format of generated files in sdp_ugprade.sh for release. | 4 months ago |
| 30907 | C. Thomas Tyler | Changed 'p4 unlock -x' to 'p4 unlock -f -x'. | 4 months ago |
| 30880 | C. Thomas Tyler |
Fixed bad grammar in comments. Non-functional change. |
4 months ago |
| 30878 | C. Thomas Tyler | Adapted ServerID detection from server.id file to use just first line. | 4 months ago |
| 30877 | C. Thomas Tyler | Refined check for SDP checkpoint counters in SDP Health Check. | 4 months ago |
| 30860 | C. Thomas Tyler |
Tweaked number of logs to keep. Fixed typo in output message. |
4 months ago |
| 30857 | C. Thomas Tyler | Clarified text in comments. | 4 months ago |
| 30856 | C. Thomas Tyler | Adapted to cleanup obsolete parallel checkpoint directories that had only *.OK files and... no *.md5 files (created by a small range of p4d versions that did not create top-level *.md5 files for parallel checkpoint directories). « |
4 months ago |
| 30855 | C. Thomas Tyler | Completed safety feature to ensure checkpoints and journals cannot never be deleted by re...move_old_checkpoints_and_journals() even if it were called at the wrong time. This is for future-proofing (and making custom SDP-derived scripts safer). « |
4 months ago |